Legacy website of the International Criminal Tribunal for the former Yugoslavia

Since the ICTY’s closure on 31 December 2017, the Mechanism maintains this website as part of its mission to preserve and promote the legacy of the UN International Criminal Tribunals.

 Visit the Mechanism's website.

Final Report to the Prosecutor by the Committee Established to Review the NATO Bombing Campaign Against the Federal Republic of Yugoslavia

printable PDF

Table of Contents

I Background and Mandate
II Review Criteria 
III Work Program
IV Assessment 


  1. General Issues 
  1. Damage to the Environment 
  2. Use of Depleted Uranium Projectiles
  3. Use of Cluster Bombs 
  4. Legal Issues Related to Target Selection 
      1. Overview of Applicable Law 
      2. Linkage Between Law Concerning  Recourse to Force and Law Concerning How Force May Be Used
      3. The Military Objective 
      4. The Principle of Proportionality 
  1. Casualty Figures 
  2. General Assessment of the Bombing Campaign
  1. Specific Incidents
    1. The Attack on a Civilian Passenger Train at the Grdelica Gorge on 12/4/99
    2. The Attack on the Djakovica Convoy on 14/4/99
    3. The Attack on the RTS (Serbian Radio and TV Station) in Belgrade on 23/4/99
    4. The Attack on the Chinese Embassyon 7/5/99
    5. The Attack on Korisa Village on 13/5/99 
  1. Recommendations 

  2. I Background and Mandate


  1. The North Atlantic Treaty Organization (NATO) conducted a bombing campaign against the Federal Republic of Yugoslavia (FRY) from 24 March 1999 to 9 June 1999. During and since that period, the Prosecutor has received numerous requests that she investigate allegations that senior political and military figures from NATO countries committed serious violations of international humanitarian law during the campaign, and that she prepares indictments pursuant to Article 18(1) & (4) of the Statute.
  2. Criticism of the NATO bombing campaign has included allegations of varying weight: a) that, as the resort to force was illegal, all NATO actions were illegal, and b) that the NATO forces deliberately attacked civilian infrastructure targets (and that such attacks were unlawful), deliberately or recklessly attacked the civilian population, and deliberately or recklessly caused excessive civilian casualties in disregard of the rule of proportionality by trying to fight a "zero casualty" war for their own side. Allegations concerning the "zero casualty" war involve suggestions that, for example, NATO aircraft operated at heights which enabled them to avoid attack by Yugoslav defences and, consequently, made it impossible for them to properly distinguish between military or civilian objects on the ground. Certain allegations went so far as to accuse NATO of crimes against humanity and genocide.
  3. Article 18 of the Tribunal’s Statute provides:

    "The Prosecutor shall initiate investigations ex officio or on the basis of information obtained from any source, particularly from Governments, United Nations organs, intergovernmental and non-governmental organizations. The Prosecutor shall assess the information received or obtained and decide whether there is sufficient basis to proceed".


  4. On 14 May 99 the then Prosecutor established a committee to assess the allegations and material accompanying them, and advise the Prosecutor and Deputy Prosecutor whether or not there is a sufficient basis to proceed with an investigation into some or all the allegations or into other incidents related to the NATO bombing.


  5. In the course of its work, the committee has not addressed in detail the issue of the fundamental legality of the use of force by NATO members against the FRY as, if such activity was unlawful, it could constitute a crime against peace and the ICTY has no jurisdiction over this offence. (See, however, paras 30 – 34 below). It is noted that the legitimacy of the recourse to force by NATO is a subject before the International Court of Justice in a case brought by the FRY against various NATO countries.
  6. II Review Criteria


  7. In the course of its review, the committee has applied the same criteria to NATO activities that the Office of the Prosecutor (OTP) has applied to the activities of other actors in the territory of the former Yugoslavia. The committee paid particular heed to the following questions:
    1. Are the prohibitions alleged sufficiently well-established as violations of international humanitarian law to form the basis of a prosecution, and does the application of the law to the particular facts reasonably suggest that a violation of these prohibitions may have occurred? and
    2. upon the reasoned evaluation of the information by the committee, is the information credible and does it tend to show that crimes within the jurisdiction of the Tribunal may have been committed by individuals during the NATO bombing campaign ?

This latter question reflects the earlier approach in relation to Article 18(1) of the Statute taken by the Prosecutor when asserting her right to investigate allegations of crimes committed by Serb forces in Kosovo (Request by the Prosecutor, Pursuant to Rule 7 bis) (B) that the President Notify the Security Council That the Federal Republic of Yugoslavia Has Failed to Comply With Its Obligations Under Article 29, dated 1 February 1999). The threshold test expressed therein by the Prosecutor was that of "credible evidence tending to show that crimes within the jurisdiction of the Tribunal may have been committed in Kosovo". That test was advanced to explain in what situation the Prosecutor would consider, for jurisdiction purposes, that she had a legal entitlement to investigate. (As a corollary, any investigation failing to meet that test could be said to be arbitrary and capricious, and to fall outside the Prosecutor’s mandate). Thus formulated, the test represents a negative cut-off point for investigations. The Prosecutor may, in her discretion require that a higher threshold be met before making a positive decision that there is sufficient basis to proceed under Article 18(1). (In fact, in relation to the situation on the ground in Kosovo, the Prosecutor was in possession of a considerable body of evidence pointing to the commission of widespread atrocities by Serb forces.) In practice, before deciding to open an investigation in any case, the Prosecutor will also take into account a number of other factors concerning the prospects for obtaining evidence sufficient to prove that the crime has been committed by an individual who merits prosecution in the international forum.


III Work Program


  1. The committee has reviewed:
    1. documents sent to the OTP by persons or groups wishing the OTP to commence investigations of leading persons from NATO countries,
    2. public documents made available by NATO, the US Department of Defense and the British Ministry of Defence,
    3. documents filed by the FRY before the ICJ, a large number of other FRY documents, and also the two volume compilation of the FRY Ministry of Foreign Affairs entitled NATO Crimes in Yugoslavia (White Book),
    4. various documents submitted by Human Rights Watch including a letter sent to the Secretary General of NATO during the bombing campaign, a paper on NATO’s Use of Cluster Munitions, and a report on Civilian Deaths in the NATO Air Campaign,
    5. a UNEP study: The Kosovo Conflict: Consequence for the Environment and Human Settlements,
    6. documents submitted by a Russian Parliamentary Commission,
    7. two studies by a German national, Mr. Ekkehard Wenz, one concerning the bombing of a train at the Grdelica Gorge and the other concerning the bombing of the Djakovica Refugee Convoy,
    8. various newspaper reports and legal articles as they have come to the attention of committee members,
    9. the response to a letter containing a number of questions sent to NATO by the OTP, and
    10. an Amnesty International Report entitled "Collateral Damage" or Unlawful Killings? Violations of the Laws of War by NATO during Operation Allied Force.
  1. It should be noted that the committee did not travel to the FRY and it did not solicit information from the FRY through official channels as no such channels existed during the period when the review was conducted. Most of the material reviewed by the committee was in the public domain. The committee has relied exclusively on documents. The FRY submitted to the Prosecutor a substantial amount of material concerning particular incidents. In attempting to assess what happened on the ground, the committee relied upon the Human Rights Watch Report entitled Civilian Deaths in the NATO Air Campaign and upon the documented accounts in the FRY Ministry of Foreign Affairs volumes entitled NATO Crimes in Yugoslavia. The committee also relied heavily on NATO press statements and on the studies done by Mr. Ekkehard Wenz. The information available was adequate for making a preliminary assessment of incidents in which civilians were killed or injured. Information related to attacks on objects where civilians were not killed or injured was difficult to obtain and very little usable information was obtained.
  2. To assist in the preparation of an Interim Report, a member of the Military Analysis Team reviewed the documents available in the OTP at the time, that is, all those referred to in paragraph 6 above except the FRY volumes entitled NATO Crimes in Yugoslavia, the HRW report on Civilian Deaths in the NATO Air Campaign, the studies by Mr. Wenz, NATO’s response to the letter sent by the OTP to NATO, and the Amnesty International Report. The analyst prepared: a) a list of key incidents, b) a list of civilian residential targets, c) a list of civilian facility targets, d) a list of cultural property targets, e) a list of power facility targets, f) a list of targets the destruction of which might significantly affect the environment, and g) a list of communications targets. Very little information was available concerning the targets in lists (b) through (g).
  3. The committee reviewed the above lists and requested the preparation of a file containing all available information on certain particular incidents, and on certain target categories. (It should be noted that the use of the terms "target" or "attack" in this report does not mean that in every case the site in question was deliberately struck by NATO. The terms are convenient shorthand for incidents in which it is alleged that particular locations were damaged in the course of the bombing campaign).

The key incidents and target categories were:


    1. the attack on a civilian passenger train at the Grdelica Gorge – 12/4/99 – 10 or more civilians killed, 15 or more injured,
    2. the attack on the Djakovica Convoy – 14/4/99 – 70-75 civilians killed, 100 or more injured,
    3. the attack on Surdulica, - 27/4/99 – 11 civilians killed, 100 or more injured,
    4. the attack on Cuprija – 8/4/99 – 1 civilian killed, 5 injured,
    5. the attack on the Cigota Medical Institute – 8/4/99 – 3 civilians killed,
    6. the attack on Hotels Baciste and Putnik – 13/4/99 – 1 civilian killed,
    7. the attacks on the Pancevo Petrochemical Complex and Fertilizer Company – 15/4/99 and 18/4/99 – no reported civilian casualties,
    8. the attack on the Nis Tobbaco Factory – 18/4/99 – no reported civilian casualties,
    9. the attack on the Djakovica Refugee Camp – 21/4/99 – 5 civilians killed, 16-19 injured,
    10. the attack on a bus at Lu`ane – 1/5/99 39 civilians killed,
    11. the attack on a bus at Pec – 3/5/99 – 17 civilians killed, 44 injured,
    12. the attack at Korisa village – 13/5/99 – 48-87 civilians killed,
    13. the attack on the Belgrade TV and Radio Station – 23/4/99 – 16 civilians killed,
    14. the attack on the Chinese Embassy in Belgrade – 7/5/99 – 3 civilians killed, 15 injured,
    15. attack on Nis City Centre and Hospital – 7/5/99 – 13 civilians killed, 60 injured,
    16. attack on Istok Prison – 21/5/99 – at least 19 civilians killed,
    17. attack on Belgrade Hospital – 20/5/99 – 3 civilians killed, several injured,
    18. attack on Surdulica Sanatorium – 30/5/99 – 23 killed, many injured,
    19. attack on journalists convoy Prizren-Brezovica Road – 31/5/99 – 1 civilian killed – 3 injured
    20. attack on Belgrade Heating Plant – 4/4/99, - 1 killed,
    21. attacks on Trade and Industry Targets.

10. On 23 July 1999, each committee member was provided with a binder including all available material. The committee members reviewed material in the binders.


11. In addition to reviewing factual information, the committee has also gathered legal materials and reviewed relevant legal issues, including the legality of the use of depleted uranium projectiles, the legality of the use of cluster munitions, whether or not the bombing campaign had an unlawfully adverse impact on the environment, and legal issues related to target selection.


12. The committee prepared an interim report on the basis of its analysis of the legal and factual material available and this was presented to the Prosecutor on 6 December 1999. At the direction of the Prosecutor, the committee then further updated the incident list and prepared a list of general questions and questions related to specific incidents. A letter enclosing the questionnaire and incident list was sent to NATO on 8 February 2000. A general reply was received on 10 May 2000.


13. It has not been possible for the committee to look at the NATO bombing campaign on a bomb by bomb basis and that was not its task. The committee has, however, reviewed public information concerning several incidents, including all the more well known incidents, with considerable care. It has also endeavored to examine, and has posed questions to NATO, concerning all other incidents in which it appears three or more civilians were killed.


In conducting its review, the committee has focused primarily on incidents in which civilian deaths were alleged and/or confirmed. The committee reviewed certain key incidents in depth for its interim report. These key incidents included 10 incidents in which 10 or more civilians were killed. The review by Human Rights Watch revealed 12 incidents in which 10 or more civilians were killed, all of the incidents identified by the committee plus two additional incidents: a) the attack on the Aleksinak "Deligrad" military barracks on 5/5/99 in which 10 civilians were killed and 30 wounded (a bomb aimed at the barracks fell short), and b) the attack on a military barracks in Novi Pazar on 31/5/99 in which 11 civilians were killed and 23 wounded (5 out of 6 munitions hit the target but one went astray). The committee’s review of incidents in which it is alleged fewer than three civilians were killed has been hampered by a lack of reliable information.


IV Assessment


  1. General Issues
  1. Damage to the Environment

14. The NATO bombing campaign did cause some damage to the environment. For instance, attacks on industrial facilities such as chemical plants and oil installations were reported to have caused the release of pollutants, although the exact extent of this is presently unknown. The basic legal provisions applicable to protection of the environment in armed conflict are Article 35(3) of Additional Protocol I, which states that ‘[i]t is prohibited to employ methods or means of warfare which are intended, or may be expected, to cause widespread, long-term and severe damage to the natural environment’ and Article 55 which states:


1. Care shall be taken in warfare to protect the natural environment against widespread, long-term and severe damage. This protection includes a prohibition of the use of methods or means of warfare which are intended or may be expected to cause such damage to the natural environment and thereby to prejudice the health or survival of the population.


2. Attacks against the natural environment by way of reprisals are prohibited


15. Neither the USA nor France has ratified Additional Protocol I. Article 55 may, nevertheless, reflect current customary law (see however the 1996 Advisory Opinion on the Legality of Nuclear Weapons, where the International Court of Justice appeared to suggest that it does not (ICJ Rep. (1996), 242, para. 31)). In any case, Articles 35(3) and 55 have a very high threshold of application. Their conditions for application are extremely stringent and their scope and contents imprecise. For instance, it is generally assumed that Articles 35(3) and 55 only cover very significant damage. The adjectives ‘widespread, long-term, and severe’ used in Additional Protocol I are joined by the word ‘and’, meaning that it is a triple, cumulative standard that needs to be fulfilled.
Consequently, it would appear extremely difficult to develop a prima facie case upon the basis of these provisions, even assuming they were applicable. For instance, it is thought that the notion of ‘long-term’ damage in Additional Protocol I would need to be measured in years rather than months, and that as such, ordinary battlefield damage of the kind caused to France in World War I would not be covered.
The great difficulty of assessing whether environmental damage exceeded the threshold of Additional Protocol I has also led to criticism by ecologists. This may partly explain the disagreement as to whether any of the damage caused by the oil spills and fires in the 1990/91 Gulf War technically crossed the threshold of Additional Protocol I.
It is the committee’s view that similar difficulties would exist in applying Additional Protocol I to the present facts, even if reliable environmental assessments were to give rise to legitimate concern concerning the impact of the NATO bombing campaign. Accordingly, these effects are best considered from the underlying principles of the law of armed conflict such as necessity and proportionality.


16. The conclusions of the Balkan Task Force (BTF) established by UNEP to look into the Kosovo situation are:


"Our findings indicate that the Kosovo conflict has not caused an environmental catastrophe affecting the Balkans region as a whole.


Nevertheless, pollution detected at some sites is serious and poses a threat to human health.


BTF was able to identify environmental ‘hot spots’, namely in Pancevo, Kragujevac, Novi Sad and Bor, where immediate action and also further monitoring and analyses will be necessary. At all of these sites, environmental contamination due to the consequences of the Kosovo conflict was identified.


Part of the contamination identified at some sites clearly pre-dates the Kosovo conflict, and there is evidence of long-term deficiencies in the treatment and storage of hazardous waste.


The problems identified require immediate attention, irrespective of their cause, if further damage to human health and the environment is to be avoided."


17. The OTP has been hampered in its assessment of the extent of environmental damage in Kosovo by a lack of alternative and corroborated sources regarding the extent of environmental contamination caused by the NATO bombing campaign. Moreover, it is quite possible that, as this campaign occurred only a year ago, the UNEP study may not be a reliable indicator of the long term environmental consequences of the NATO bombing, as accurate assessments regarding the long-term effects of this contamination may not yet be practicable.


It is the opinion of the committee, on the basis of information currently in its possession, that the environmental damage caused during the NATO bombing campaign does not reach the Additional Protocol I threshold. In addition, the UNEP Report also suggests that much of the environmental contamination which is discernible cannot unambiguously be attributed to the NATO bombing.


18. The alleged environmental effects of the NATO bombing campaign flow in many cases from NATO’s striking of legitimate military targets compatible with Article 52 of Additional Protocol I such as stores of fuel, industries of fundamental importance for the conduct of war and for the manufacture of supplies and material of a military character, factories or plant and manufacturing centres of fundamental importance for the conduct of war. Even when targeting admittedly legitimate military objectives, there is a need to avoid excessive long-term damage to the economic infrastructure and natural environment with a consequential adverse effect on the civilian population. Indeed, military objectives should not be targeted if the attack is likely to cause collateral environmental damage which would be excessive in relation to the direct military advantage which the attack is expected to produce (A.P.V. Rogers, "Zero Casualty Warfare," IRRC, March 2000, Vol. 82, pp. 177-8).


19. It is difficult to assess the relative values to be assigned to the military advantage gained and harm to the natural environment, and the application of the principle of proportionality is more easily stated than applied in practice. In applying this principle, it is necessary to assess the importance of the target in relation to the incidental damage expected: if the target is sufficiently important, a greater degree of risk to the environment may be justified.


20. The adverse effect of the coalition air campaign in the Gulf war upon the civilian infrastructure prompted concern on the part of some experts regarding the notion of "military objective." This has prompted some experts to argue that where the presumptive effect of hostilities upon the civilian infrastructure (and consequently the civilian population) is grave, the military advantage conferred by the destruction of the military objective would need to be decisive (see below, paras. 40–41). Similar considerations would, in the committee’s view, be warranted where the grave threat to the civilian infrastructure emanated instead from excessive environmental harm resulting from the hostilities. The critical question is what kind of environmental damage can be considered to be excessive. Unfortunately, the customary rule of proportionality does not include any concrete guidelines to this effect.


21. The military worth of the target would need to be considered in relation to the circumstances prevailing at the time. If there is a choice of weapons or methods of attack available, a commander should select those which are most likely to avoid, or at least minimize, incidental damage. In doing so, however, he is entitled to take account of factors such as stocks of different weapons and likely future demands, the timeliness of attack and risks to his own forces (A.P.V. Rogers, ibid, at p. 178). Operational reality is recognized in the Statute of the International Criminal Court, an authoritative indicator of evolving customary international law on this point, where Article 8(b)(iv) makes the infliction of incidental environmental damage an offence only if the attack is launched intentionally in the knowledge that it will cause widespread, long-term and severe damage to the natural environment which would be clearly excessive in relation to the concrete and direct overall military advantage anticipated. The use of the word "clearly’ ensures that criminal responsibility would be entailed only in cases where the excessiveness of the incidental damage was obvious.


22. Taken together, this suggests that in order to satisfy the requirement of proportionality, attacks against military targets which are known or can reasonably be assumed to cause grave environmental harm may need to confer a very substantial military advantage in order to be considered legitimate. At a minimum, actions resulting in massive environmental destruction, especially where they do not serve a clear and important military purpose, would be questionable. The targeting by NATO of Serbian petro-chemical industries may well have served a clear and important military purpose.


23. The above considerations also suggest that the requisite mens rea on the part of a commander would be actual or constructive knowledge as to the grave environmental effects of a military attack; a standard which would be difficult to establish for the purposes of prosecution and which may provide an insufficient basis to prosecute military commanders inflicting environmental harm in the (mistaken) belief that such conduct was warranted by military necessity. (In the Hostages case before the Nuremberg Military Tribunals, for instance, the German General Rendulic was acquitted of the charge of wanton devastation on the grounds that although Rendulic may have erred in believing that there was military necessity for the widespread environmental destruction entailed by his use of a ‘scorched earth’ policy in the Norwegian province of Finnmark, he was not guilty of a criminal act (11 Trials of War Criminals, (1950), 1296)). In addition, the notion of ‘excessive’ environmental destruction is imprecise and the actual environmental impact, both present and long term, of the NATO bombing campaign is at present unknown and difficult to measure.


24. In order to fully evaluate such matters, it would be necessary to know the extent of the knowledge possessed by NATO as to the nature of Serbian military-industrial targets (and thus, the likelihood of environmental damage flowing from their destruction), the extent to which NATO could reasonably have anticipated such environmental damage (for instance, could NATO have reasonably expected that toxic chemicals of the sort allegedly released into the environment by the bombing campaign would be stored alongside that military target?) and whether NATO could reasonably have resorted to other (and less environmentally damaging) methods for achieving its military objective of disabling the Serbian military-industrial infrastructure.


25. It is therefore the opinion of the committee, based on information currently available to it, that the OTP should not commence an investigation into the collateral environmental damage caused by the NATO bombing campaign.


  1. Use of Depleted Uranium Projectiles

26. There is evidence of use of depleted uranium (DU) projectiles by NATO aircraft during the bombing campaign. There is no specific treaty ban on the use of DU projectiles. There is a developing scientific debate and concern expressed regarding the impact of the use of such projectiles and it is possible that, in future, there will be a consensus view in international legal circles that use of such projectiles violate general principles of the law applicable to use of weapons in armed conflict. No such consensus exists at present. Indeed, even in the case of nuclear warheads and other weapons of mass-destruction – those which are universally acknowledged to have the most deleterious environmental consequences – it is difficult to argue that the prohibition of their use is in all cases absolute. (Legality of Nuclear Weapons, ICJ Rep. (1996), 242). In view of the uncertain state of development of the legal standards governing this area, it should be emphasised that the use of depleted uranium or other potentially hazardous substance by any adversary to conflicts within the former Yugoslavia since 1991 has not formed the basis of any charge laid by the Prosecutor. It is acknowledged that the underlying principles of the law of armed conflict such as proportionality are applicable also in this context; however, it is the committee’s view that analysis undertaken above (paras. 14-25) with regard to environmental damage would apply, mutatis mutandis, to the use of depleted uranium projectiles by NATO. It is therefore the opinion of the committee, based on information available at present, that the OTP should not commence an investigation into use of depleted uranium projectiles by NATO.


  1. Use of Cluster Bombs

27. Cluster bombs were used by NATO forces during the bombing campaign. There is no specific treaty provision which prohibits or restricts the use of cluster bombs although, of course, cluster bombs must be used in compliance with the general principles applicable to the use of all weapons. Human Rights Watch has condemned the use of cluster bombs alleging that the high "dud" or failure rate of the submunitions (bomblets) contained inside cluster bombs converts these submunitions into antipersonnel landmines which, it asserts, are now prohibited under customary international law. Whether antipersonnel landmines are prohibited under current customary law is debatable, although there is a strong trend in that direction. There is, however, no general legal consensus that cluster bombs are, in legal terms, equivalent to antipersonnel landmines. It should be noted that the use of cluster bombs was an issue of sorts in the Martić Rule 61 Hearing Decision of Trial Chamber I on 8 March 1996. In that decision the Chamber stated there was no formal provision forbidding the use of cluster bombs as such (para. 18 of judgment) but it regarded the use of the Orkan rocket with a cluster bomb warhead in that particular case as evidence of the intent of the accused to deliberately attack the civilian population because the rocket was inaccurate, it landed in an area with no military objectives nearby, it was used as an antipersonnel weapon launched against the city of Zagreb and the accused indicated he intended to attack the city as such (paras. 23-31 of judgment). The Chamber concluded that "the use of the Orkan rocket in this case was not designed to hit military targets but to terrorise the civilians of Zagreb" (para. 31 of judgment). There is no indication cluster bombs were used in such a fashion by NATO. It is the opinion of the committee, based on information presently available, that the OTP should not commence an investigation into use of cluster bombs as such by NATO.


  1. Legal Issues Related to Target Selection
  1. Overview of Applicable Law

28. In brief, in combat military commanders are required: a) to direct their operations against military objectives, and b) when directing their operations against military objectives, to ensure that the losses to the civilian population and the damage to civilian property are not disproportionate to the concrete and direct military advantage anticipated. Attacks which are not directed against military objectives (particularly attacks directed against the civilian population) and attacks which cause disproportionate civilian casualties or civilian property damage may constitute the actus reus for the offence of unlawful attack under Article 3 of the ICTY Statute. The mens rea for the offence is intention or recklessness, not simple negligence. In determining whether or not the mens rea requirement has been met, it should be borne in mind that commanders deciding on an attack have duties:


a) to do everything practicable to verify that the objectives to be attacked are military objectives,


b) to take all practicable precautions in the choice of methods and means of warfare with a view to avoiding or, in any event to minimizing incidental civilian casualties or civilian property damage, and


c)to refrain from launching attacks which may be expected to cause disproportionate civilian casualties or civilian property damage.


29. One of the principles underlying international humanitarian law is the principle of distinction, which obligates military commanders to distinguish between military objectives and civilian persons or objects. The practical application of this principle is effectively encapsulated in Article 57 of Additional Protocol which, in part, obligates those who plan or decide upon an attack to "do everything feasible to verify that the objectives to be attacked are neither civilians nor civilian objects". The obligation to do everything feasible is high but not absolute. A military commander must set up an effective intelligence gathering system to collect and evaluate information concerning potential targets. The commander must also direct his forces to use available technical means to properly identify targets during operations. Both the commander and the aircrew actually engaged in operations must have some range of discretion to determine which available resources shall be used and how they shall be used. Further, a determination that inadequate efforts have been made to distinguish between military objectives and civilians or civilian objects should not necessarily focus exclusively on a specific incident. If precautionary measures have worked adequately in a very high percentage of cases then the fact they have not worked well in a small number of cases does not necessarily mean they are generally inadequate.


b) Linkage Between Law Concerning Recourse to Force and Law Concerning How Force May Be Used


30. Allegations have been made that, as NATO’s resort to force was not authorized by the Security Council or in self-defence, that the resort to force was illegal and, consequently, all forceful measures taken by NATO were unlawful. These allegations justify a brief discussion of the jus ad bellum. In brief, the jus ad bellum regulates when states may use force and is, for the most part, enshrined in the UN Charter. In general, states may use force in self defence (individual or collective) and for very few other purposes. In particular, the legitimacy of the presumed basis for the NATO bombing campaign, humanitarian intervention without prior Security Council authorization, is hotly debated. That being said, as noted in paragraph 4 above, the crime related to an unlawful decision to use force is the crime against peace or aggression. While a person convicted of a crime against peace may, potentially, be held criminally responsible for all of the activities causing death, injury or destruction during a conflict, the ICTY does not have jurisdiction over crimes against peace.


31. The jus in bello regulates how states may use force. The ICTY has jurisdiction over serious violations of international humanitarian law as specified in Articles 2-5 of the Statute. These are jus in bello offences.


32. The precise linkage between jus ad bellum and jus in bello is not completely resolved. There were suggestions by the prosecution before the International Military Tribunal at Nuremberg and in some other post World War II war crimes cases that all of the killing and destruction caused by German forces were war crimes because the Germans were conducting an aggressive war. The courts were unreceptive to these arguments. Similarly, in the 1950’s there was a debate concerning whether UN authorized forces were required to comply with the jus in bello as they represented the good side in a battle between good an evil. This debate died out as the participants realized that a certain crude reciprocity was essential if the law was to have any positive impact. An argument that the "bad" side had to comply with the law while the "good" side could violate it at will would be most unlikely to reduce human suffering in conflict.


33. More recently, a refined approach to the linkage issue has been advocated by certain law of war scholars. Using their approach, assuming that the only lawful basis for recourse to force is self defence, each use of force during a conflict must be measured by whether or not it complies with the jus in bello and by whether or not it complies with the necessity and proportionality requirements of self defence. The difficulty with this approach is that it does not adequately address what should be done when it is unclear who is acting in self defence and it does not clarify the obligations of the "bad" side.


34. As a matter of practice, which we consider to be in accord with the most widely accepted and reputable legal opinion, we in the OTP have deliberately refrained from assessing jus ad bellum issues in our work and focused exclusively on whether or not individuals have committed serious violations of international humanitarian law as assessed within the confines of the jus in bello.


  1. The military objective

35. The most widely accepted definition of "military objective" is that in Article 52 of Additional Protocol I which states in part:

In so far as objects are concerned, military objectives are limited to those objects which by their nature, location, purpose or use make an effective contribution to military action and whose total or partial destruction, capture or neutralization, in the circumstances ruling at the time, offers a definite military advantage.


36. Where objects are concerned, the definition has two elements: (a) their nature, location, purpose or use must make an effective contribution to military action, and (b) their total or partial destruction, capture or neutralization must offer a definite military advantage in the circumstances ruling at the time. Although this definition does not refer to persons, in general, members of the armed forces are considered combatants, who have the right to participate directly in hostilities, and as a corollary, may also be attacked.


37. The definition is supposed to provide a means whereby informed objective observers (and decision makers in a conflict) can determine whether or not a particular object constitutes a military objective. It accomplishes this purpose in simple cases. Everyone will agree that a munitions factory is a military objective and an unoccupied church is a civilian object. When the definition is applied to dual-use objects which have some civilian uses and some actual or potential military use (communications systems, transportation systems, petrochemical complexes, manufacturing plants of some types), opinions may differ. The application of the definition to particular objects may also differ depending on the scope and objectives of the conflict. Further, the scope and objectives of the conflict may change during the conflict.


38. Using the Protocol I definition and his own review of state practice, Major General A.P.V. Rogers, a former Director of British Army Legal Services has advanced a tentative list of military objectives:


military personnel and persons who take part in the fighting without being members of the armed forces, military facilities, military equipment, including military vehicles, weapons, munitions and stores of fuel, military works, including defensive works and fortifications, military depots and establishments, including War and Supply Ministries, works producing or developing military supplies and other supplies of military value, including metallurgical, engineering and chemical industries supporting the war effort; areas of land of military significance such as hills, defiles and bridgeheads; railways, ports, airfields, bridges, main roads as well as tunnels and canals; oil and other power installations; communications installations, including broadcasting and television stations and telephone and telegraph stations used for military communications. (Rogers, Law on the Battlefield (1996) 37)


The list was not intended to be exhaustive. It remains a requirement that both elements of the definition must be met before a target can be properly considered an appropriate military objective.


39. In 1956, the International Committee of the Red Cross (ICRC) drew up the following proposed list of categories of military objectives:


I. The objectives belonging to the following categories are those considered to be of generally recognized military importance:


(1) Armed forces, including auxiliary or complementary organisations, and persons who, though not belonging to the above-mentioned formations, nevertheless take part in the fighting.


(2) Positions, installations or constructions occupied by the forces indicated in sub-paragraph 1 above, as well as combat objectives (that is to say, those objectives which are directly contested in battle between land or sea forces including airborne forces).


(3) Installations, constructions and other works of a military nature, such as barracks, fortifications, War Ministries (e.g. Ministries of Army, Navy, Air Force, National Defence, Supply) and other organs for the direction and administration of military operations.


(4) Stores of army or military supplies, such as munition dumps, stores of equipment or fuel, vehicles parks.


(5) Airfields, rocket launching ramps and naval base installations.


(6) Those of the lines and means of communications (railway lines, roads, bridges, tunnels and canals) which are of fundamental military importance.


(7) The installations of broadcasting and television stations; telephone and telegraph exchanges of fundamental military importance.


(8) Industries of fundamental importance for the conduct of the war:


(a) industries for the manufacture of armaments such as weapons, munitions, rockets, armoured vehicles, military aircraft, fighting ships, including the manufacture of accessories and all other war material;


(b) industries for the manufacture of supplies and material of a military character, such as transport and communications material, equipment of the armed forces;


(c) factories or plant constituting other production and manufacturing centres of fundamental importance for the conduct of war, such as the metallurgical, engineering and chemical industries, whose nature or purpose is essentially military;


(d) storage and transport installations whose basic function it is to serve the industries referred to in (a)-(c);


(e) installations providing energy mainly for national defence, e.g. coal, other fuels, or atomic energy, and plants producing gas or electricity mainly for military consumption.


(9) Installations constituting experimental, research centres for experiments on and the development of weapons and war material.


II. The following however, are excepted from the foregoing list:


(1) Persons, constructions, installations or transports which are protected under the Geneva Conventions I, II, III, of August 12, 1949;


(2) Non-combatants in the armed forces who obviously take no active or direct part in hostilities.


III. The above list will be reviewed at intervals of not more than ten years by a group of Experts composed of persons with a sound grasp of military strategy and of others concerned with the protection of the civilian population.
(Y. Sandoz, C. Swiniarski, B. Zimmerman, eds., Commentary on the Additional Protocols of 8 June 1977 to the Geneva Conventions of 12 August 1949 (1987) at 632-633.


40. The Protocol I definition of military objective has been criticized by W. Hays Parks, the Special Assistant for Law of War Matters to the U.S. Army Judge Advocate General as being focused too narrowly on definite military advantage and paying too little heed to war sustaining capability, including economic targets such as export industries. (W. Hays Parks, "Air War and the Law of War," 32 A.F.L. Rev. 1, 135-45 (1990)). On the other hand, some critics of Coalition conduct in the Gulf War have suggested that the Coalition air campaign, directed admittedly against legitimate military objectives within the scope of the Protocol I definition, caused excessive long-term damage to the Iraqi economic infrastructure with a consequential adverse effect on the civilian population. (Middle East Watch, Needless Deaths in the Gulf War: Civilian Casualties during the Air Campaign and Violations of the Laws of War (1991); Judith G. Gardam, "Proportionality and Force in International Law," 87 Am. J. Int'l L. 391, 404-10 (1993)).


41. This criticism has not gone unexplored. Françoise Hampson, a British scholar, has suggested a possible refinement of the definition:


In order to determine whether there is a real subject of concern here, it would be necessary to establish exactly what the effect has been of the damage to the civilian infrastructure brought about by the hostilities. If that points to a need further to refine the law, it is submitted that what is needed is a qualification to the definition of military objectives. Either it should require the likely cumulative effect on the civilian population of attacks against such targets to be taken into account, or the same result might be achieved by requiring that the destruction of the object offer a definite military advantage in the context of the war aim. Françoise Hampson, "Means and Methods of Warfare in the Conflict in the Gulf," in P. Rowe, ed., The Gulf War 1990-91 in International and English Law 89 (1983) 100.


42. Although the Protocol I definition of military objective is not beyond criticism, it provides the contemporary standard which must be used when attempting to determine the lawfulness of particular attacks. That being said, it must be noted once again neither the USA nor France is a party to Additional Protocol I. The definition is, however, generally accepted as part of customary law.


43. To put the NATO campaign in context, it is instructive to look briefly at the approach to the military objective concept in history of air warfare. The Protocol I standard was not applicable during World War II. The bomber offensives conducted during that war were conducted with technological means which rendered attacks on targets occupying small areas almost impossible. In general, depending upon the period in the conflict, bomber attacks could be relied upon, at best, to strike within 5 miles, 2 miles or 1 mile of the designated target. The mission for the US/UK Combined Bomber Offensive from the UK was:


    "To conduct a joint United States-British air offensive to accomplish the progressive destruction and dislocation of the German military, industrial and economic system, and the undermining of the morale of the German people to a point where their capacity for armed resistance is fatally weakened. This is construed as meaning so weakened as to permit initiation of final combined operations on the Continent."


    (A. Verrier, The Bomber Offensive (1968) 330).


    The principal specific objectives of the offensive were designated as:


    "Submarine construction yards and bases.
    German aircraft industry.
    Ball bearings.
    Synthetic rubber and tires.
    Military transport vehicles."
    (A. Verrier, ibid, at 330).


    Notwithstanding the designation of specific targets and the attempt, at least by US Army Air Force commanders on occasion, to conduct a precision bombing campaign, for the most part World War II bombing campaigns were aimed at area targets and intended, directly or indirectly, to affect the morale of the enemy civilian population. It is difficult to describe the fire bombing of Hamburg, Dresden and Tokyo as anything other than attacks intended to kill, terrorize or demoralize civilians. Whether or not these attacks could be justified legally in the total war context of the time, they would be unlawful if they were required to comply with Protocol I.


44. Technology, law, and the public consensus of what was acceptable, at least in demonstrably limited conflicts, had evolved by the time of the 1990-91 Gulf Conflict. Technological developments, such as precision guided munitions, and the rapid acquisition of control of the aerospace by coalition air forces significantly enhanced the precision with which targets could be attacked.


Target sets used during the Gulf Conflict were:


    "Leadership; Command, Control, and Communications; Strategic Air Defenses; Airfields; Nuclear, Biological, and Chemical Research and Production; Naval Forces and Port Facilities; Military Storage and Production; Railroads and Bridges, Electrical Power; and Oil Refining and Distribution Facilities. Schwarzkopf added the Republican Guard as a category and Scuds soon emerged as a separate target set. After the beginning of Desert Storm, two more categories appeared: fixed surface-to-air missile sites in the KTO and breaching sites for the ground offensive."


    (W. Murray, Air War in the Persian Gulf (1995) 32)


45. In the words of the Cohen, Shelton Joint Statement on Kosovo given to the US Senate:


"At the outset of the air campaign, NATO set specific strategic objectives for its use of force in Kosovo that later served as the basis for its stated conditions to Milosevic for stopping the bombing. These objectives were to:


-- Demonstrate the seriousness of NATO’s opposition to Belgrade’s aggression in the Balkans;


-- Deter Milosevic from continuing and escalating his attacks on helpless civilians and create conditions to reverse his ethnic cleansing; and


-- Damage Serbia’s capacity to wage war against Kosovo in the future or spread the war to neighbors by diminishing or degrading its ability to wage military operations...


Phases of the Campaign. Operation Allied Force was originally planned to be prosecuted in five phases under NATO’s operational plan, the development of which began in the summer of 1998. Phase 0 was the deployment of air assets into the European theater. Phase 1 would establish air superiority over Kosovo and degrade command and control over the whole of the FRY. Phase 2 would attack military targets in Kosovo and those FRY forces south of 44 degrees north latitude, which were providing reinforcement to Serbian forces into Kosovo. This was to allow targeting of forces not only in Kosovo, but also in the FRY south of Belgrade. Phase 3 would expand air operations against a wide range of high-value military and security force targets throughout the FRY. Phase 4 would redeploy forces as required. A limited air response relying predominantly on cruise missiles to strike selected targets throughout the Phase 1. Within a few days of the start of NATO’s campaign, alliance aircraft were striking both strategic and tactical targets throughout Serbia, as well as working to suppress and disrupt the FRY’s integrated air defence system.


At the NATO Summit in Washington on April 23, 1999, alliance leaders decided to further intensify the air campaign by expanding the target set to include military-industrial infrastructure, media, and other strategic targets ...."


46. The NATO Internet Report Kosovo One Year On (http://www.nato.int/kosovo/repo 2000, 21 Mar 00) described the targets as:


"The air campaign set out to weaken Serb military capabilities, both strategically and tactically. Strikes on tactical targets, such as artillery and field headquarters, had a more immediate effect in disrupting the ethnic cleansing of Kosovo. Strikes against strategic targets, such as government ministries and refineries, had long term and broader impact on the Serb military machine.


The bulk of NATO’s effort against tactical targets was aimed at military facilities , fielded forces, heavy weapons, and military vehicles and formations in Kosovo and southern Serbia...


Strategic targets included Serb air defences, command and control facilities, Yugoslav military (VJ) and police (MUP) forces headquarters, and supply routes".


47. Most of the targets referred to in the quotations above are clearly military objectives. The precise scope of "military-industrial infrastructure, media and other strategic targets" as referred to in the US statement and "government ministries and refineries" as referred to in the NATO statement is unclear. Whether the media constitutes a legitimate target group is a debatable issue. If the media is used to incite crimes, as in Rwanda, then it is a legitimate target. If it is merely disseminating propaganda to generate support for the war effort, it is not a legitimate target.


d) The Principle of Proportionality


48. The main problem with the principle of proportionality is not whether or not it exists but what it means and how it is to be applied. It is relatively simple to state that there must be an acceptable relation between the legitimate destructive effect and undesirable collateral effects. For example, bombing a refugee camp is obviously prohibited if its only military significance is that people in the camp are knitting socks for soldiers. Conversely, an air strike on an ammunition dump should not be prohibited merely because a farmer is plowing a field in the area. Unfortunately, most applications of the principle of proportionality are not quite so clear cut. It is much easier to formulate the principle of proportionality in general terms than it is to apply it to a particular set of circumstances because the comparison is often between unlike quantities and values. One cannot easily assess the value of innocent human lives as opposed to capturing a particular military objective.


49. The questions which remain unresolved once one decides to apply the principle of proportionality include the following:


a) What are the relative values to be assigned to the military advantage gained and the injury to non-combatants and or the damage to civilian objects?


b) What do you include or exclude in totaling your sums?


c) What is the standard of measurement in time or space? and


d) To what extent is a military commander obligated to expose his own forces to danger in order to limit civilian casualties or damage to civilian objects?


50. The answers to these questions are not simple. It may be necessary to resolve them on a case by case basis, and the answers may differ depending on the background and values of the decision maker. It is unlikely that a human rights lawyer and an experienced combat commander would assign the same relative values to military advantage and to injury to noncombatants. Further, it is unlikely that military commanders with different doctrinal backgrounds and differing degrees of combat experience or national military histories would always agree in close cases. It is suggested that the determination of relative values must be that of the "reasonable military commander". Although there will be room for argument in close cases, there will be many cases where reasonable military commanders will agree that the injury to noncombatants or the damage to civilian objects was clearly disproportionate to the military advantage gained.


51. Much of the material submitted to the OTP consisted of reports that civilians had been killed, often inviting the conclusion to be drawn that crimes had therefore been committed. Collateral casualties to civilians and collateral damage to civilian objects can occur for a variety of reasons. Despite an obligation to avoid locating military objectives within or near densely populated areas, to remove civilians from the vicinity of military objectives, and to protect their civilians from the dangers of military operations, very little prevention may be feasible in many cases. Today’s technological society has given rise to many dual use facilities and resources. City planners rarely pay heed to the possibility of future warfare. Military objectives are often located in densely populated areas and fighting occasionally occurs in such areas. Civilians present within or near military objectives must, however, be taken into account in the proportionality equation even if a party to the conflict has failed to exercise its obligation to remove them.


52. In the Kupreskic Judgment (Case No: IT-95-16-T 14 Jan 2000) the Trial Chamber addressed the issue of proportionality as follows:


"526. As an example of the way in which the Martens clause may be utilised, regard might be had to considerations such as the cumulative effect of attacks on military objectives causing incidental damage to civilians. In other words, it may happen that single attacks on military objectives causing incidental damage to civilians, although they may raise doubts as to their lawfulness, nevertheless do not appear on their face to fall foul per se of the loose prescriptions of Articles 57 and 58 (or of the corresponding customary rules). However, in case of repeated attacks, all or most of them falling within the grey area between indisputable legality and unlawfulness, it might be warranted to conclude that the cumulative effect of such acts entails that they may not be in keeping with international law. Indeed, this pattern of military conduct may turn out to jeopardise excessively the lives and assets of civilians, contrary to the demands of humanity."


This formulation in Kupreskic can be regarded as a progressive statement of the applicable law with regard to the obligation to protect civilians. Its practical import, however, is somewhat ambiguous and its application far from clear. It is the committee’s view that where individual (and legitimate) attacks on military objectives are concerned, the mere cumulation of such instances, all of which are deemed to have been lawful, cannot ipso facto be said to amount to a crime. The committee understands the above formulation, instead, to refer to an overall assessment of the totality of civilian victims as against the goals of the military campaign.


V Casualty Figures


53. In its report, Civilian Deaths in the NATO Air Campaign, Human Rights Watch documented some 500 civilian deaths in 90 separate incidents. It concluded: "on the basis available on these ninety incidents that as few as 488 and as many as 527 Yugoslav civilians were killed as a result of NATO bombing. Between 62 and 66 percent of the total registered civilian deaths occurred in just twelve incidents. These twelve incidents accounted for 303 to 352 civilian deaths. These were the only incidents among the ninety documented in which ten or more civilian deaths were confirmed." Ten of these twelve incidents were included among the incidents which were reviewed with considerable care by the committee (see para. 9 above) and our estimate was that between 273 and 317 civilians were killed in these ten incidents. Human Rights Watch also found the FRY Ministry of Foreign Affairs publication NATO Crimes in Yugoslavia to be largely credible on the basis of its own filed research and correlation with other sources. A review of this publication indicates it provides an estimated total of approximately 495 civilians killed and 820 civilians wounded in specific documented instances. For the purposes of this report, the committee operates on the basis of the number of persons allegedly killed as found in both publications. It appears that a figure similar to both publications would be in the range of 500 civilians killed.


VI General Assesment of the Bombing Campaign


54. During the bombing campaign, NATO aircraft flew 38,400 sorties, including 10,484 strike sorties. During these sorties, 23, 614 air munitions were released (figures from NATO). As indicated in the preceding paragraph, it appears that approximately 500 civilians were killed during the campaign. These figures do not indicate that NATO may have conducted a campaign aimed at causing substantial civilian casualties either directly or incidentally.


55. The choice of targets by NATO (see paras. 38 and 39 above) includes some loosely defined categories such as military-industrial infrastructure and government ministries and some potential problem categories such as media and refineries. All targets must meet the criteria for military objectives (see para. 28-30 above). If they do not do so, they are unlawful. A general label is insufficient. The targeted components of the military-industrial infrastructure and of government ministries must make an effective contribution to military action and their total or partial destruction must offer a definite military advantage in the circumstances ruling at the time. Refineries are certainly traditional military objectives but tradition is not enough and due regard must be paid to environmental damage if they are attacked (see paras. 14-25 above). The media as such is not a traditional target category. To the extent particular media components are part of the C3 (command, control and communications) network they are military objectives. If media components are not part of the C3 network then they may become military objectives depending upon their use. As a bottom line, civilians, civilian objects and civilian morale as such are not legitimate military objectives. The media does have an effect on civilian morale. If that effect is merely to foster support for the war effort, the media is not a legitimate military objective. If the media is used to incite crimes, as in Rwanda, it can become a legitimate military objective. If the media is the nerve system that keeps a war-monger in power and thus perpetuates the war effort, it may fall within the definition of a legitimate military objective. As a general statement, in the particular incidents reviewed by the committee, it is the view of the committee that NATO was attempting to attack objects it perceived to be legitimate military objectives.


56. The committee agrees there is nothing inherently unlawful about flying above the height which can be reached by enemy air defences. However, NATO air commanders have a duty to take practicable measures to distinguish military objectives from civilians or civilian objectives. The 15,000 feet minimum altitude adopted for part of the campaign may have meant the target could not be verified with the naked eye. However, it appears that with the use of modern technology, the obligation to distinguish was effectively carried out in the vast majority of cases during the bombing campaign.


  1. Specific Incidents

57. In the course of its review, the committee did not come across any incident which, in its opinion, required investigation by the OTP. The five specific incidents discussed below are those which, in the opinion of the committee, were the most problematic. The facts cited in the discussion of each specific incident are those indicated in the information within the possession of the OTP at the time of its review.


  1. The Attack on a Civilian Passenger Train at the Grdelica Gorge on 12/4/99

58. On 12 April 1999, a NATO aircraft launched two laser guided bombs at the Leskovac railway bridge over the Grdelica gorge and Juzna Morava river, in eastern Serbia. A 5-carriage passenger train, travelling from Belgrade to Ristovac on the Macedonian border, was crossing the bridge at the time, and was struck by both missiles. The various reports made of this incident concur that the incident occurred at about 11.40 a.m. At least ten people were killed in this incident and at least 15 individuals were injured. The designated target was the railway bridge, which was claimed to be part of a re-supply route being used for Serb forces in Kosovo. After launching the first bomb, the person controlling the weapon, at the last instant before impact, sighted movement on the bridge. The controller was unable to dump the bomb at that stage and it hit the train, the impact of the bomb cutting the second of the passenger coaches in half. Realising the bridge was still intact, the controller picked a second aim point on the bridge at the opposite end from where the train had come and launched the second bomb. In the meantime the train had slid forward as a result of the original impact and parts of the train were also hit by the second bomb.


59. It does not appear that the train was targeted deliberately. US Deputy Defense Secretary John Hamre stated that "one of our electro-optically guided bombs homed in on a railroad bridge just when a passenger train raced to the aim point. We never wanted to destroy that train or kill its occupants. We did want to destroy the bridge and we regret this accident." The substantive part of the explanation, both for the failure to detect the approach of the passenger train and for firing a second missile once it had been hit by the first, was given by General Wesley Clark, NATO’s Supreme Allied Commander for Europe and is here reprinted in full:


"[T]his was a case where a pilot was assigned to strike a railroad bridge that is part of the integrated communications supply network in Serbia. He launched his missile from his aircraft that was many miles away, he was not able to put his eyes on the bridge, it was a remotely directed attack. And as he stared intently at the desired target point on the bridge, and I talked to the team at Aviano who was directly engaged in this operation, as the pilot stared intently at the desired aim point on the bridge and worked it, and worked it and worked it, and all of a sudden at the very last instant with less than a second to go he caught a flash of movement that came into the screen and it was the train coming in.


Unfortunately he couldn’t dump the bomb at that point, it was locked, it was going into the target and it was an unfortunate incident which he, and the crew, and all of us very much regret. We certainly don’t want to do collateral damage.


The mission was to take out the bridge. He realised when it had happened that he had not hit the bridge, but what he had hit was the train. He had another aim point on the bridge, it was a relatively long bridge and he believed he still had to accomplish his mission, the pilot circled back around. He put his aim point on the other end of the bridge from where the train had come, by the time the bomb got close to the bridge it was covered with smoke and clouds and at the last minute again in an uncanny accident, the train had slid forward from the original impact and parts of the train had moved across the bridge, and so that by striking the other end of the bridge he actually caused additional damage to the train." (Press Conference, NATO HQ, Brussels, 13 April).


General Clark then showed the cockpit video of the plane which fired on the bridge:


"The pilot in the aircraft is looking at about a 5-inch screen, he is seeing about this much and in here you can see this is the railroad bridge which is a much better view than he actually had, you can see the tracks running this way.


Look very intently at the aim point, concentrate right there and you can see how, if you were focused right on your job as a pilot, suddenly that train appeared. It was really unfortunate.


Here, he came back around to try to strike a different point on the bridge because he was trying to do a job to take the bridge down. Look at this aim point – you can see smoke and other obscuration there – he couldn’t tell what this was exactly.


Focus intently right at the centre of the cross. He is bringing these two crosses together and suddenly he recognises at the very last instant that the train that was struck here has moved on across the bridge and so the engine apparently was struck by the second bomb." (Press Conference, NATO HQ, Brussels, 13 April).


60. Some doubt has since been cast on this version of events by a comprehensive technical report submitted by a German national, Mr Ekkehard Wenz, which queries the actual speed at which the events took place in relation to that suggested by the video footage of the incident released by NATO. The effect of this report is to suggest that the reaction time available to the person controlling the bombs was in fact considerably greater than that alleged by NATO. Mr. Wenz also suggests the aircraft involved was an F15E Strike Eagle with a crew of two and with the weapons being controlled by a Weapons Systems Officer (WSO) not the pilot.


61. The committee has reviewed both the material provided by NATO and the report of Mr. Wenz with considerable care. It is the opinion of the committee that it is irrelevant whether the person controlling the bomb was the pilot or the WSO. Either person would have been travelling in a high speed aircraft and likely performing several tasks simultaneously, including endeavouring to keep the aircraft in the air and safe from surrounding threats in a combat environment. If the committee accepts Mr. Wenz’s estimate of the reaction time available, the person controlling the bombs still had a very short period of time, less than 7 or 8 seconds in all probability, to react. Although Mr Wenz is of the view that the WSO intentionally targeted the train, the committee’s review of the frames used in the report indicates another interpretation is equally available. The cross hairs remain fixed on the bridge throughout, and it is clear from this footage that the train can be seen moving toward the bridge only as the bomb is in flight: it is only in the course of the bomb’s trajectory that the image of the train becomes visible. At a point where the bomb is within a few seconds of impact, a very slight change to the bomb aiming point can be observed, in that it drops a couple of feet. This sequence regarding the bomb sights indicates that it is unlikely that the WSO was targeting the train, but instead suggests that the target was a point on the span of the bridge before the train appeared.


62. It is the opinion of the committee that the bridge was a legitimate military objective. The passenger train was not deliberately targeted. The person controlling the bombs, pilot or WSO, targeted the bridge and, over a very short period of time, failed to recognize the arrival of the train while the first bomb was in flight. The train was on the bridge when the bridge was targeted a second time and the bridge length has been estimated at 50 meters (Wenz study para 6 g above at p.25). It is the opinion of the committee that the information in relation to the attack with the first bomb does not provide a sufficient basis to initiate an investigation. The committee has divided views concerning the attack with the second bomb in relation to whether there was an element of recklessness in the conduct of the pilot or WSO. Despite this, the committee is in agreement that, based on the criteria for initiating an investigation (see para. 5 above), this incident should not be investigated. In relation to whether there is information warranting consideration of command responsibility, the committee is of the view that there is no information from which to conclude that an investigation is necessary into the criminal responsibility of persons higher in the chain of command. Based on the information available to it, it is the opinion of the committee that the attack on the train at Grdelica Gorge should not be investigated by the OTP.


(ii) The Attack on the Djakovica Convoy on 14/4/99


63. The precise facts concerning this incident are difficult to determine. In particular, there is some confusion about the number of aircraft involved, the number of bombs dropped, and whether one or two convoys were attacked. The FRY Ministry of Foreign Affairs Report (White Book) describes the incident as follows:


"On April 14, 1999 […] on the Djakovica-Prizren road, near the villages of Madanaj and Meja, a convoy of Albanian refugees was targeted three times. Mostly women, children and old people were in the convoy, returning to their homes in cars, on tractors and carts. The first assault on the column of over 1000 people took place while they were moving through Meja village. Twelve persons were killed on that occasion. The people from the convoy scattered around and tried to find shelter in the nearby houses. But NATO warplanes launched missiles on those houses as well, killing another 7 persons in the process. The attack continued along the road beween [the] villages [of] Meja and Bistrazin. One tractor with trailer was completely destroyed. Twenty people out of several of them on the tractor were killed. In the repeated attack on the refugee vehicles, one more person was killed." (Vol 1, p.1)


Total casualty figures seem to converge around 70-75 killed with approximately 100 injured. The FRY publication NATO War Crimes in Yugoslavia states 73 were killed and 36 were wounded.


64. NATO initially denied, but later acknowledged, responsibility for this attack. Assuming the facts most appropriate to a successful prosecution, NATO aircraft flying at 15000 feet or higher to avoid Yugoslav air defences attacked two vehicle convoys, both of which contained civilian vehicles. On 15 April, NATO confirmed that the aircraft had been flying at an altitude of 15,000 feet (approximately 5 km) and that, in this attack, the pilots had viewed the target with the naked eye rather than remotely. The aim of the attack was to destroy Serb military forces, in the area of Djakovica, who had been seen by NATO aircraft setting fire to civilian houses. At a Press Conference of 15 April 1999, NATO claimed that this was an area where the Yugoslav Special Police Forces, the MUP, were conducting ethnic cleansing operations over the preceding days. The road between Prizren and Djakovica served as an important resupply and reinforcement route for the Yugoslav Army and the Special Police.


65. A reconstruction of what is known about the attack reveals that in the hours immediately prior to the attack, at around 1030, NATO forces claimed to have seen a progression of burning villages, and that a series of fires could be seen progressing to the south east. They formed the view that MUP and VJ forces were thus methodically working from the north to the south through villages, setting them ablaze and forcing all the Kosovar Albanians out of those villages. At around 1030, the pilot spotted a three-vehicle convoy near to the freshest burning house, and saw uniformly shaped dark green vehicles which appeared to be troop carrying vehicles. He thus formed the view that the convoy comprised VJ and MUP forces working their way down towards Djakovica and that they were preparing to set the next house on fire. In response, an F-16 bombed the convoy’s lead vehicle at approximately 1110; the pilot relayed a threat update and the coordinates of the attack and departed the area to refuel. A second F-16 aircraft appears to have arrived on the scene around 1135, and visually assessed the target area as containing large vehicles which were located near a complex of buildings. A single GBU-12 bomb was dropped at 1148. Contemporaneously, a third aircraft identified a large convoy on a major road south east out of Djakovica and sought to identify the target. The target was verified as a VJ convoy at 1216 and an unspecified number of bombs were dropped at 1219. In the next 15 or so minutes (exact time unspecified), the same aircraft appears to have destroyed one further vehicle in the convoy. Simultaneously, two Jaguar aircraft each dropped 1 GBU-12 bomb each, but both missed their targets. Between 1235 and 1245, the first F-16 aircraft appears to have dropped three further bombs, at least one of which appears to have missed its target.


66. It is claimed by one source (report on file with the OTP) that the Yugoslav TV broadcast of the attack on the Djakovica convoy on 15 April 1999 recorded a conversation between one F-16 pilot involved in the attack and the AWACs. This conversation is alleged to establish both that the attack on the convoy was deliberate and that a UK Harrier pilot had advised the F-16 pilot that the convoy was comprised solely of tractors and civilians. The F-16 pilot was then allegedly told that the convoy was nevertheless a legitimate military target and was instructed to fire on it. This same report also suggests that the convoy was attacked with cluster bombs, indicated by bomb remnants and craters left at the site. However, these claims – both with regard to the foreknowledge of the pilot as to the civilian nature of the convoy and of the weapons used – are not confirmed by any other source.


67. NATO itself claimed that although the cockpit video showed the vehicles to look like tractors, when viewed with the naked eye from the attack altitude they appeared to be military vehicles. They alleged that several characteristics indicated it to be a military convoy including movement, size, shape, colour, spacing and high speed prior to the attack. There had also been reports of Serb forces using civilian vehicles. An analysis of the Serb TV footage of the attack on Djakovica by the OTP indicates that at approximately 1240, some point during the attack, doubt was conveyed that Serb convoys do not usually travel in convoys of that size. However, the on-scene analysis of the convoy appeared to convey the impression that the convoy comprised a mix of military and civilian vehicles. At around 1300, an order appears to have been issued, suspending attacks until the target could be verified.


68. NATO has consistently claimed that it believed the Djakovica convoy to be escorted by Serb military vehicles at the time of the attack. Human Rights Watch has commented on the incident as follows:


"General Clark stated in September that NATO consistently observed Yugoslav military vehicles moving on roads "intermixed with civilian convoys." After the Djakovica-Decane incident, General Clark says, "we got to be very, very cautious about striking objects moving on the roads. Another NATO officer, Col. Ed Boyle, says: "Because we were so concerned with collateral damage, the CFAC [Combined Forces Air Component Commander] at the time, General [Michael] Short, put out the guidance that if military vehicles were intermingled with civilian vehicles, they were not to be attacked, due to the collateral damage." When this directive was actually issued remains an important question. Nevertheless, the change in NATO rules of engagement indicates that the alliance recognized that it had taken insufficient precautions in mounting this attack, in not identifying civilians present, and in assuming that the intended targets were legitimate military objectives rather than in positively identifying them."


69. It is the opinion of the committee that civilians were not deliberately attacked in this incident. While there is nothing unlawful about operating at a height above Yugoslav air defences, it is difficult for any aircrew operating an aircraft flying at several hundred miles an hour and at a substantial height to distinguish between military and civilian vehicles in a convoy. In this case, most of the attacking aircraft were F16s with a crew of one person to fly the aircraft and identify the target. As soon as the crews of the attacking aircraft became aware of the presence of civilians, the attack ceased.


70. While this incident is one where it appears the aircrews could have benefitted from lower altitude scrutiny of the target at an early stage, the committee is of the opinion that neither the aircrew nor their commanders displayed the degree of recklessness in failing to take precautionary measures which would sustain criminal charges. The committee also notes that the attack was suspended as soon as the presence of civilians in the convoy was suspected. Based on the information assessed, the committee recommends that the OTP not commence an investigation related to the Djakovica Convoy bombing.


iii) The Bombing of the RTS (Serbian TV and Radio Station) in Belgrade on 23/4/99


71. On 23 April 1999, at 0220, NATO intentionally bombed the central studio of the RTS (state-owned) broadcasting corporation at 1 Aberdareva Street in the centre of Belgrade. The missiles hit the entrance area, which caved in at the place where the Aberdareva Street building was connected to the Takovska Street building. While there is some doubt over exact casualty figures, between 10 and 17 people are estimated to have been killed.


72. The bombing of the TV studio was part of a planned attack aimed at disrupting and degrading the C3 (Command, Control and Communications) network. In co-ordinated attacks, on the same night, radio relay buildings and towers were hit along with electrical power transformer stations. At a press conference on 27 April 1999, NATO officials justified this attack in terms of the dual military and civilian use to which the FRY communication system was routinely put, describing this as a


"very hardened and redundant command and control communications system [which …] uses commercial telephone, […] military cable, […] fibre optic cable, […] high frequency radio communication, […] microwave communication and everything can be interconnected. There are literally dozens, more than 100 radio relay sites around the country, and […] everything is wired in through dual use. Most of the commercial system serves the military and the military system can be put to use for the commercial system […]."


Accordingly, NATO stressed the dual-use to which such communications systems were put, describing civilian television as "heavily dependent on the military command and control system and military traffic is also routed through the civilian system" (press conference of 27 April, ibid).


73. At an earlier press conference on 23 April 1999, NATO officials reported that the TV building also housed a large multi-purpose communications satellite antenna dish, and that "radio relay control buildings and towers were targeted in the ongoing campaign to degrade the FRY’s command, control and communications network". In a communication of 17 April 1999 to Amnesty International, NATO claimed that the RTS facilities were being used "as radio relay stations and transmitters to support the activities of the FRY military and special police forces, and therefore they represent legitimate military targets" (Amnesty International Report, NATO/Federal Republic of Yugoslavia: Violations of the Laws of War by NATO during Operation Allied Force, June 2000, p. 42).


74. Of the electrical power transformer stations targeted, one transformer station supplied power to the air defence co-ordination network while the other supplied power to the northern-sector operations centre. Both these facilities were key control elements in the FRY integrated air-defence system. In this regard, NATO indicated that


"we are not targeting the Serb people as we repeatedly have stated nor do we target President Milosevic personally, we are attacking the control system that is used to manipulate the military and security forces."


More controversially, however, the bombing was also justified on the basis of the propaganda purpose to which it was employed:


"[We need to] directly strike at the very central nerve system of Milosovic’s regime. This of course are those assets which are used to plan and direct and to create the political environment of tolerance in Yugoslavia in which these brutalities can not only be accepted but even condoned. [….] Strikes against TV transmitters and broadcast facilities are part of our campaign to dismantle the FRY propaganda machinery which is a vital part of President Milosevic’s control mechanism."


In a similar statement, British Prime Minister Tony Blair was reported as saying in The Times that the media "is the apparatus that keeps him [Milosević] in power and we are entirely justified as NATO allies in damaging and taking on those targets” (24 April, 1999). In a statement of 8 April 1999, NATO also indicated that the TV studios would be targeted unless they broadcast 6 hours per day of Western media reports: "If President Milosevic would provide equal time for Western news broadcasts in its programmes without censorship 3 hours a day between noon and 1800 and 3 hours a day between 1800 and midnight, then his TV could be an acceptable instrument of public information."


75. NATO intentionally bombed the Radio and TV station and the persons killed or injured were civilians. The questions are: was the station a legitimate military objective and; if it was, were the civilian casualties disproportionate to the military advantage gained by the attack? For the station to be a military objective within the definition in Article 52 of Protocol I: a) its nature, purpose or use must make an effective contribution to military action and b) its total or partial destruction must offer a definite military advantage in the circumstances ruling at the time. The 1956 ICRC list of military objectives, drafted before the Additional Protocols, included the installations of broadcasting and television stations of fundamental military importance as military objectives (para. 39 above). The list prepared by Major General Rogers included broadcasting and television stations if they meet the military objective criteria (para. 38 above). As indicated in paras. 72 and 73 above, the attack appears to have been justified by NATO as part of a more general attack aimed at disrupting the FRY Command, Control and Communications network, the nerve centre and apparatus that keeps Milosević in power, and also as an attempt to dismantle the FRY propaganda machinery. Insofar as the attack actually was aimed at disrupting the communications network, it was legally acceptable.


76. If, however, the attack was made because equal time was not provided for Western news broadcasts, that is, because the station was part of the propaganda machinery, the legal basis was more debatable. Disrupting government propaganda may help to undermine the morale of the population and the armed forces, but justifying an attack on a civilian facility on such grounds alone may not meet the "effective contribution to military action" and "definite military advantage" criteria required by the Additional Protocols (see paras. 35-36, above). The ICRC Commentary on the Additional Protocols interprets the expression "definite military advantage anticipated" to exclude "an attack which only offers potential or indeterminate advantages" and interprets the expression "concrete and direct" as intended to show that the advantage concerned should be substantial and relatively close rather than hardly perceptible and likely to appear only in the long term (ICRC Commentary on the Additional Protocols of 8 June 1977, para. 2209). While stopping such propaganda may serve to demoralize the Yugoslav population and undermine the government’s political support, it is unlikely that either of these purposes would offer the "concrete and direct" military advantage necessary to make them a legitimate military objective. NATO believed that Yugoslav broadcast facilities were "used entirely to incite hatred and propaganda" and alleged that the Yugoslav government had put all private TV and radio stations in Serbia under military control (NATO press conferences of 28 and 30 April1999). However, it was not claimed that they were being used to incite violence akin to Radio Milles Collines during the Rwandan genocide, which might have justified their destruction (see para. 47 above). At worst, the Yugoslav government was using the broadcasting networks to issue propaganda supportive of its war effort: a circumstance which does not, in and of itself, amount to a war crime (see in this regard the judgment of the International Military Tribunal in Nuremberg in 1946 in the case of Hans Fritzsche, who served as a senior official in the Propaganda ministry alleged to have incited and encouraged the commission of crimes. The IMT held that although Fritzsche clearly made strong statements of a propagandistic nature, it was nevertheless not prepared to find that they were intended to incite the commission of atrocities, but rather, were aimed at arousing popular sentiment in support of Hitler and the German war effort (American Journal of International Law, vol. 41 (1947) 328)). The committee finds that if the attack on the RTS was justified by reference to its propaganda purpose alone, its legality might well be questioned by some experts in the field of international humanitarian law. It appears, however, that NATO’s targeting of the RTS building for propaganda purposes was an incidental (albeit complementary) aim of its primary goal of disabling the Serbian military command and control system and to destroy the nerve system and apparatus that keeps Milosević in power. In a press conference of 9 April 1999, NATO declared that TV transmitters were not targeted directly but that "in Yugoslavia military radio relay stations are often combined with TV transmitters [so] we attack the military target. If there is damage to the TV transmitters, it is a secondary effect but it is not [our] primary intention to do that." A NATO spokesperson, Jamie Shea, also wrote to the Brussels-based International Federation of Journalists on 12 April claiming that OperationAllied Force "target[ed] military targets only and television and radio towers are only struck if they [were] integrated into military facilities … There is no policy to strike television and radio transmitters as such" (cited in Amnesty International Report, ibid, June 2000).


77. Assuming the station was a legitimate objective, the civilian casualties were unfortunately high but do not appear to be clearly disproportionate.


Although NATO alleged that it made "every possible effort to avoid civilian casualties and collateral damage" (Amnesty International Report, ibid, June 2000, p. 42), some doubts have been expressed as to the specificity of the warning given to civilians by NATO of its intended strike, and whether the notice would have constituted "effective warning … of attacks which may affect the civililan population, unless circumstances do not permit" as required by Article 57(2) of Additional Protocol I.


Evidence on this point is somewhat contradictory. On the one hand, NATO officials in Brussels are alleged to have told Amnesty International that they did not give a specific warning as it would have endangered the pilots (Amnesty International Report, ibid, June 2000, at p. 47; see also para. 49 above re: proportionality and the extent to which a military commander is obligated to expose his own forces to danger in order to limit civilian casualties or damage). On this view, it is possible that casualties among civilians working at the RTS may have been heightened because of NATO’s apparent failure to provide clear advance warning of the attack, as required by Article 57(2).


On the other hand, foreign media representatives were apparently forewarned of the attack (Amnesty International Report, ibid). As Western journalists were reportedly warned by their employers to stay away from the television station before the attack, it would also appear that some Yugoslav officials may have expected that the building was about to be struck. Consequently, UK Prime Minister Tony Blair blamed Yugoslav officials for not evacuating the building, claiming that "[t]hey could have moved those people out of the building. They knew it was a target and they didn’t … [I]t was probably for … very clear propaganda reasons." (ibid, citing Moral combat – NATO at war, broadcast on BBC2 on 12 March 2000). Although knowledge on the part of Yugoslav officials of the impending attack would not divest NATO of its obligation to forewarn civilians under Article 57(2), it may nevertheless imply that the Yugoslav authorities may be partially responsible for the civilian casualties resulting from the attack and may suggest that the advance notice given by NATO may have in fact been sufficient under the circumstances.


78. Assuming the RTS building to be a legitimate military target, it appeared that NATO realised that attacking the RTS building would only interrupt broadcasting for a brief period. Indeed, broadcasting allegedly recommenced within hours of the strike, thus raising the issue of the importance of the military advantage gained by the attack vis-à-vis the civilian casualties incurred. The FRY command and control network was alleged by NATO to comprise a complex web and that could thus not be disabled in one strike. As noted by General Wesley Clark, NATO "knew when we struck that there would be alternate means of getting the Serb Television. There’s no single switch to turn off everything but we thought it was a good move to strike it and the political leadership agreed with us" (ibid, citing "Moral combat, NATO at War," broadcast on BBC2 on 12 March 2000). At a press conference on 27 April 1999, another NATO spokesperson similarly described the dual-use Yugoslav command and control network as "incapable of being dealt with in "a single knock-out blow (ibid)." The proportionality or otherwise of an attack should not necessarily focus exclusively on a specific incident. (See in this regard para. 52, above, referring to the need for an overall assessment of the totality of civilian victims as against the goals of the military campaign). With regard to these goals, the strategic target of these attacks was the Yugoslav command and control network. The attack on the RTS building must therefore be seen as forming part of an integrated attack against numerous objects, including transmission towers and control buildings of the Yugoslav radio relay network which were "essential to Milosevic’s ability to direct and control the repressive activities of his army and special police forces in Kosovo" (NATO press release, 1 May 1999) and which comprised "a key element in theYugoslav air-defence network" (ibid, 1 May1999). Attacks were also aimed at electricity grids that fed the command and control structures of the Yugoslav Army (ibid, 3 May 1999). Other strategic targets included additional command and control assets such as the radio and TV relay sites at Novi Pazar, Kosovaka and Krusevac (ibid) and command posts (ibid, 30 April). Of the electrical power transformer stations targeted, one transformer station supplied power to the air-defence coordination network while the other supplied power to the northern sector operations centre. Both these facilities were key control elements in the FRY integrated air-defence system (ibid, 23 April 1999). The radio relay and TV transmitting station near Novi Sad was also an important link in the air defence command and control communications network. Not only were these targets central to the Federal Republic of Yugoslavia’s governing apparatus, but formed, from a military point of view, an integral part of the strategic communications network which enabled both the military and national command authorities to direct the repression and atrocities taking place in Kosovo (ibid, 21 April 1999).


79. On the basis of the above analysis and on the information currently available to it, the committee recommends that the OTP not commence an investigation related to the bombing of the Serbian TV and Radio Station.


  1. The Attack on the Chinese Embassy on 7/5/99

80. On 7/5/99, at 2350, NATO aircraft fired several missiles which hit the Chinese Embassy in Belgrade, killing 3 Chinese citizens, injuring an estimated 15 others, and causing extensive damage to the embassy building and other buildings in the immediate surrounds. At the moment of the attack, fifty people were reported to have been in the embassy buildings. By the admission of US Government sources, the Chinese Embassy compound was mistakenly hit. The bombing occurred because at no stage in the process was it realised that the bombs were aimed at the Chinese Embassy. The Embassy had been wrongly identified as the Yugoslav Federal Directorate for Supply and Procurement (Yugoimport FDSP) at 2 Umetnosti Boulevard in New Belgrade. The FDSP was deemed by the CIA to be a legitimate target due to its role in military procurement: it was selected for its role in support of the Yugoslav military effort.


81. Under Secretary of State Thomas Pickering offered the following explanation for what occurred:


    "The bombing resulted from three basic failures. First, the technique used to locate the intended target – the headquarters of the Yugoslav Federal Directorate for Supply and Procurement (FDSP) – was severely flawed. Second, none of the military or intelligence databases used to verify target information contained the correct location of the Chinese Embassy. Third, nowhere in the target review process was either of the first two mistakes detected. No one who might have known that the targeted building was not the FDSP headquarters – but was in fact the Chinese Embassy – was ever consulted."

According to US Government sources, the street address of the intended target, the FDSP headquarters was known as Bulevar Umetnosti 2 in New Belgrade. During a mid-April "work-up" of the target to prepare a mission folder for the B-2 bomber crew, three maps were used in an attempt to physically locate this address within the neighborhood: two local commercial maps from 1989 and 1996, and one US government (National Imagery and Mapping Agency or NIMA) map produced in 1997. None of these maps had any reference to the FDSP building and none accurately identified the current location of the Chinese Embassy.


82. The root of the failures in target location appears to stem from the land navigation techniques employed by an intelligence officer in an effort to pinpoint the location of the FDSP building at Bulevar Umetnosti 2. The officer used techniques known as "intersection" and "resection" which, while appropriate to locate distant or inaccessible points or objects, are inappropriate for use in aerial targeting as they provide only an approximate location. Using this process, the individual mistakenly determined that the building which we now know to be the Chinese Embassy was the FDSP headquarters. This method of identification was not questioned or reviewed and hence this flaw in the address location process went undetected by all the others who evaluated the FDSP headquarters as a military target. It also appears that very late in the process, an intelligence officer serendipitously came to suspect that the target had been wrongly identified and sought to raise the concern that the building had been mislocated. However, throughout a series of missed opportunities, the problem of identification was not brought to the attention of the senior managers who may have been able to intervene in time to prevent the strike.


83. Finally, reviewing elements in, inter alia, the Joint Staff did not uncover either the inaccurate location of the FDSP headquarters or the correct location of the Chinese Embassy. The data base reviews were limited to validating the target data sheet geographic coordinates and the information put into the data base by the NIMA analyst. Such a circular process did not serve to uncover the original error and highlighted the system’s susceptibility to a single point of data base failure. The critical linchpin for both the error in identification of the building and the failure of the review mechanisms was thus the inadequacy of the supporting data bases and the mistaken assumption the information they contained would necessarily be accurate.


84. The building hit was clearly a civilian object and not a legitimate military objective. NATO, and subsequently various organs of the US Government, including the CIA, issued a formal apology, accepted full responsibility for the incident and asserted that the intended target, the Federal Directorate for Supply and Procurement, would have been a legitimate military objective. The USA has formally apologized to the Chinese Government and agreed to pay $28 million in compensation to the Chinese Government and $4.5 million to the families of those killed or injured. The CIA has also dismissed one intelligence officer and reprimanded six senior managers. The US Government also claims to have taken corrective actions in order to assign individual responsibility and to prevent mistakes such as this from occurring in the future.


85. It is the opinion of the committee that the aircrew involved in the attack should not be assigned any responsibility for the fact they were given the wrong target and that it is inappropriate to attempt to assign criminal responsibility for the incident to senior leaders because they were provided with wrong information by officials of another agency. Based on the information available to it, the committee is of the opinion that the OTP should not undertake an investigation concerning the bombing of the Chinese Embassy.


  1. The Attack on Korisa Village on 13/5/99

86. On 14 May 1999, NATO aircraft dropped 10 bombs on the village of Korisa, on the highway between Prizren and Pristina. Much confusion seems to exist about this incident, and factual accounts do not seem to easily tally with each other. As many as 87 civilians, mainly refugees, were killed in this attack and approximately 60 appear to have been wounded. The primary target in this attack was asserted by NATO to be a Serbian military camp and Command Post which were located near the village of Korisa. It appears that the refugees were near the attacked object. However, unlike previous cases where NATO subsequently claimed that an error had occurred in its targeting or its military intelligence sources, NATO spokespersons continued to affirm the legitimacy of this particular attack. They maintained that this was a legitimate military target and that NATO intelligence had identified a military camp and Command Post near to the village of Korisa.


87. According to NATO officials, immediately prior to the attack, the target was identified as having military revetments. The pilot was able to see silhouettes of vehicles on the ground as the attack took place at 2330, when two laser guided bombs were dropped. Ten minutes later, another two laser guided bombs and six gravity bombs were dropped. In a press conference on 15 May, NATO stated that the attack went ahead because the target was confirmed by prior intelligence as being valid and the pilot identified vehicles present. There were never any doubts, from NATO spokespersons, as to the validity of this target.


88. Information about NATO’s position on the bombardment of Koriša was released at the press conference on the following day, 15 May. At this conference, General Jertz twice affirmed that the target was, in NATO’s opinion, legitimate since military facilities were present at the site:


"As already has been mentioned, it was a legitimate military target. NATO reconnaissance and intelligence orders identified just outside Koriša a military camp and command post, including an armoured personnel carrier and 10 pieces of artillery. Follow-up intelligence confirmed this information as being a valid military target. Immediately prior to the attack at 23.30-11.30 pm – local time Thursday night an airborne forward air controller identified the target, so the identification and attack system of his aircraft, having positively identified the target as what looked like dug-in military reveted positions, he dropped two laser guided bombs. Approximately 10 minutes later, the third aircraft engaged the target with gravity bombs, with six gravity bombs. A total of 10 bombs were dropped on the target."


When questioned about the presence of civilians on the ground, General Jertz indicated:


"What I can say so far is when the pilot attacked the target he had to visually identify it through the attack systems which are in the aircraft, and you know it was by night, so he did see silhouettes of vehicles on the ground and as it was by prior intelligence a valid target, he did do the attack […] it was a legitimate target. Since late April we knew there were command posts, military pieces in that area and they have been continuously used. So for the pilot flying the attack, it was a legitimate target. But when he is in the target area for attacking, it is his responsibility to make sure that all the cues he sees are the ones which he needs to really attack. And at night he saw the silhouettes of vehicles and that is why he was allowed to attack. Of course, and we have to be very fair, we are talking at night. If there is anybody sleeping somewhere in a house, you would not be able to see it from the perspective of a pilot. But once again, don’t misinterpret it. It was a military target which had been used since the beginning of conflict over there and we have all sources used to identify this target in order to make sure that this target was still a valid target when it was attacked." (Emphasis added).


The NATO position thus appears to be that it bombed a legitimate military target, that it knew nothing of the presence of civilians and that none were observed immediately prior to the attack. Indeed, NATO stated that they believed this area to have been completely cleared of civilians. There is some information indicating that displaced Kosovar civilians were forcibly concentrated within a military camp in the village of Koriša as human shields and that Yugoslav military forces may thus be at least partially responsible for the deaths there.


89. The available information concerning this incident is in conflict. The attack occurred in the middle of the night at about 2330. The stated object of the attack was a legitimate military objective. According to NATO, all practicable precautions were taken and it was determined civilians were not present. It appears that a relatively large number of civilians were killed. It also appears these civilians were either returning refugees or persons gathered as human shields by FRY authorities or both. The committee is of the view that the credible information available is not sufficient to tend to show that a crime within the jurisdiction of the Tribunal has been committed by the aircrew or by superiors in the NATO chain of command. Based on the information available to it, the committee is of the opinion that OTP should not undertake an investigation concerning the bombing at Koriša.


V Recommendations


90. The committee has conducted its review relying essentially upon public documents, including statements made by NATO and NATO countries at press conferences and public documents produced by the FRY. It has tended to assume that the NATO and NATO countries’ press statements are generally reliable and that explanations have been honestly given. The committee must note, however, that when the OTP requested NATO to answer specific questions about specific incidents, the NATO reply was couched in general terms and failed to address the specific incidents. The committee has not spoken to those involved in directing or carrying out the bombing campaign. The committee has also assigned substantial weight to the factual assertions made by Human Rights Watch as its investigators did spend a limited amount of time on the ground in the FRY. Further, the committee has noted that Human Rights Watch found the two volume compilation of the FRY Ministry of Foreign Affairs entitled NATO Crimes in Yugoslavia generally reliable and the committee has tended to rely on the casualty figures for specific incidents in this compilation. If one accepts the figures in this compilation of approximately 495 civilians killed and 820 civilians wounded in documented instances, there is simply no evidence of the necessary crime base for charges of genocide or crimes against humanity. Further, in the particular incidents reviewed by the committee with particular care (see paras. 9, and 48-76) the committee has not assessed any particular incidents as justifying the commencement of an investigation by the OTP. NATO has admitted that mistakes did occur during the bombing campaign; errors of judgment may also have occurred. Selection of certain objectives for attack may be subject to legal debate. On the basis of the information reviewed, however, the committee is of the opinion that neither an in-depth investigation related to the bombing campaign as a whole nor investigations related to specific incidents are justified. In all cases, either the law is not sufficiently clear or investigations are unlikely to result in the acquisition of sufficient evidence to substantiate charges against high level accused or against lower accused for particularly heinous offences.


91. On the basis of information available, the committee recommends that no investigation be commenced by the OTP in relation to the NATO bombing campaign or incidents occurring during the campaign.