Rambouillet 1999: Račak, Appendix B, and the Road to NATO’s Bombing Campaign

From October to March: a ceasefire that did not resolve the conflict

By autumn 1998 the Kosovo conflict had become serious enough for the UN Security Council to adopt several resolutions. Resolution 1199 of 23 September demanded an immediate cessation of hostilities by all parties and additional steps by the FRY toward a political settlement and an end to excessive security operations.[1] In October an agreement was reached between Belgrade and international mediators. The OSCE Kosovo Verification Mission — KVM was deployed. It was not an occupying force. It was an unarmed international verification mission intended to monitor compliance with the ceasefire and agreed reductions of forces.[2] Resolution 1203 of 24 October demanded full implementation of the verification agreements.[2] The basic conflict, however, remained unresolved.

The KLA did not disarm. Yugoslav and Serbian forces did not permanently withdraw to a level that removed the dispute. Attacks, kidnappings, retaliatory operations, and troop movements continued. The ceasefire reduced the intensity of the war. It did not remove its causes. The KLA also used the ceasefire to expand.

For the Kosovo-Albanian side, the presence of international monitors reduced the danger of large-scale state operations in some areas. That had an unintended consequence. The KLA gained room to reorganize, recruit, return to some areas, and expand its political influence. Western documents themselves acknowledged violations on both sides. In January 1999 NATO publicly warned that its decisions on military action would also take account of the actions of the Kosovo-Albanian leadership and all armed elements, and it was considering steps against weapons smuggling to the KLA.[3] That matters. At the start of 1999, the West still did not treat the KLA as a fully disciplined army of an already legitimate state. It was an armed actor that was also being pressed to respect the ceasefire.

Serbian and Yugoslav forces also violated the October arrangements. Serbian police and Yugoslav military forces continued to operate against KLA-held or contested areas. Later ICTY findings described continuing non-compliance with parts of the October arrangements and ongoing preparations by Serbian and Yugoslav security structures for operations in Kosovo.[4] The situation at the end of 1998 was therefore not peace suddenly shattered by one incident. It was an unstable ceasefire between parties that were still preparing for renewed conflict.

8–12 January 1999: Serbian police were killed before Račak

The Račak story should not begin only with the bodies seen on 16 January. On 8 January 1999, three Serbian police officers were killed in a KLA attack on a police patrol between Suva Reka and Štimlje. Another officer died on 12 January from wounds suffered two days earlier.[4] The ICTY later found that the operation at Račak occurred in the context of those attacks and the attempt to arrest people believed to be responsible.[4] This is important context. Račak was not a village without a KLA presence, and the police operation did not occur without prior armed violence. But that context alone does not tell us what happened during the operation itself.

January: a joint MUP–VJ operation. Early on 15 January 1999, an operation began in the village of Račak/Reçak. In the later trial of Vlastimir Đorđević, the ICTY found that the operation involved both Serbian Interior Ministry forces and the Yugoslav Army.[4] Military forces were positioned around the area, while police units entered the village and conducted searches.[4] The KLA was present in the village. That prevents a simplistic description of Račak as a politically and militarily neutral place. The same judgment, however, found the other side of the story. At least 45 Kosovo Albanians were killed. Roughly 20 to 24 bodies bore head wounds and indications of firing from close range. Among the dead were also a woman and a child.[4]

What did Belgrade claim?

Yugoslav and Serbian authorities said the action was a lawful anti-terrorist operation against the KLA. According to their account, most of the dead were KLA fighters killed in combat. Public claims also emerged that bodies had later been moved or dressed to create the appearance of a massacre of civilians. Because Račak became a major political turning point before Rambouillet, the question is legitimate: did the international reaction follow an actual killing of civilians or a staged event? What did the Finnish forensic team find?

The European Union sent a forensic team led by Finnish pathologist Helena Ranta. The investigation was not ideal. The team could begin its work only around a week after the event, which limited some conclusions.[5] Even so, its findings were important: at least 40 examined victims were unarmed civilians; there were no signs that the bodies had been systematically altered after death to manufacture evidence; and gunshot-residue testing did not establish that the dead had been shooters.[5] Forensics alone cannot reconstruct every minute of an operation. But the central claim that a large number of fighters had been systematically redressed and staged as civilians was not supported by that forensic investigation.

What did the ICTY later find? This requires a distinction between two different proceedings. In the Milutinović et al. trial, Račak was not finally litigated as a separate charge because the Trial Chamber restricted evidence on that specific allegation for procedural reasons.[6] It is therefore inaccurate to say: “The Milutinović judgment proved the Račak massacre.” It did not do so in that form. In the later Đorđević trial, however, the Chamber examined Račak in detail.[4] It found: that the action was a joint VJ–MUP operation;

that at least 45 Kosovo Albanians were killed; that the force used was highly disproportionate to the asserted arrest or anti-terrorist objective; and that the police later arranged or presented the scene in a way designed to give a false impression of what had occurred.[4] That last point matters. The court did not find that the KLA staged the bodies to manufacture a massacre. It instead found subsequent police manipulation of the scene to portray the operation as a legitimate anti-terrorist action.[4]

William Walker: the diplomat who turned the event into a global crisis. KVM head William Walker went to Račak on 16 January. Members of the mission saw a group of bodies and other dead in the village. Walker almost immediately condemned the event publicly as a massacre of civilians and blamed FRY/Serbian forces. Belgrade declared him persona non grata, although he was not in practice removed.[7] Walker’s rapid public conclusion became controversial. Critics argue that he made a political judgment before the forensic investigation was complete. That is a legitimate methodological criticism. The forensic report came later.

But the later forensic and judicial record did not validate the Serbian claim that the scene had been broadly fabricated using dead fighters. The distinction is therefore: Walker made a very rapid political characterization. That is different from asking: did later evidence overturn the central proposition that civilians had been killed? It did not.

Račak was a political turning point, not the only cause of war

After Račak, Western diplomatic pressure increased dramatically. But the statement: “NATO bombed Yugoslavia because of Račak” leaves out too much. Before Račak there had already been: the 1998 war; hundreds of deaths; large-scale displacement; Security Council Resolutions 1199 and 1203; the prior threat of NATO air strikes; the October agreement; the KVM; and continuing violations. Račak changed the political tempo. It did not create the entire crisis.

Rambouillet: negotiations under threat of force

Talks opened at the Château de Rambouillet near Paris on 6 February 1999. They were led by the Contact Group. On one side was the FRY/Serbian delegation. On the other was the Kosovo-Albanian delegation, including representatives associated with Ibrahim Rugova, other political figures, and the KLA. Before the talks even began, NATO on 30 January authorized its Secretary General to approve air strikes if the process failed.[3] Rambouillet was therefore not a conventional negotiation free of external coercive pressure. One side faced an explicit threat of military force. The Kosovo-Albanian side was also told that it could face political consequences if it rejected the proposal.[3]

What did the political settlement offer Kosovo?

The proposal would have created exceptionally broad self-government for Kosovo for a three-year interim period. Kosovo would have had its own political institutions, president, assembly, government, judiciary, extensive authority over internal affairs, local policing, minority protections, and international supervision of elections and implementation.[8] Most Serbian police would have withdrawn. The role of the Yugoslav Army inside Kosovo would have been tightly limited. The KLA, however, would also have been required to demilitarize.[8][9] At the same time, the agreement formally reaffirmed the sovereignty and territorial integrity of the Federal Republic of Yugoslavia.[8] Rambouillet was therefore not a document granting Kosovo immediate independence. Did Rambouillet guarantee an independence referendum after three years?

This is another frequently simplified claim. The text itself did not explicitly say: “after three years there shall be a binding referendum on independence.” Chapter 8 provided for an international meeting after three years to determine a mechanism for a final settlement on the basis of the will of the people, the views of relevant authorities, implementation of the agreement, and the Helsinki Final Act.[8] The later British parliamentary inquiry therefore concluded that the agreement itself did not contain an explicit promise of a binding referendum.[9] There was, however, an additional American political dimension. Madeleine Albright sent the Kosovo-Albanian delegation a letter interpreting the wording as supporting a referendum on Kosovo’s final status after three years.[9] Two readings therefore coexisted.

For Belgrade: the interim arrangement could become a path toward secession. For the Kosovo-Albanian side: without a credible route toward a referendum, the agreement did not go far enough. The ambiguity was part of what made a compromise possible at all. The KLA did not immediately want to sign either.

Rambouillet is now often described simply as: the Albanians accepted; the Serbs refused. That became true at the end. It was not yet the full picture on 23 February. The Kosovo-Albanian delegation was initially unwilling to sign the whole text. A major issue was the requirement that the KLA demilitarize.[9] Hashim Thaçi and others needed time to secure support from KLA commanders. At the end of the first phase, the Kosovo-Albanian delegation accepted the political framework in principle and left for consultations. The Serbian/Yugoslav delegation also spoke of accepting or continuing negotiations on the political element, but did not accept the military implementation arrangements and NATO deployment.[10] February therefore did not yet produce a simple signed agreement by one side and rejection by the other.

Appendix B: the part of Rambouillet that cannot be skipped

The most controversial element of the agreement was not the autonomy package itself. It was Appendix B, part of the military chapter. It defined the status of NATO-led forces. Under its terms NATO personnel would have enjoyed: immunity from FRY civil, administrative, and criminal jurisdiction; protection from arrest and detention; free entry and exit; use of transport infrastructure; and, together with vehicles, vessels, aircraft, and equipment, free and unrestricted passage and unimpeded access throughout the FRY, including associated airspace and territorial waters.[8] The language also permitted movement, bivouac, billeting, use of facilities, and logistics needed for the mission.[8] This is not a theory about a secret clause. It is part of the actual draft text.

Why does Appendix B matter? Because the military mission was not restricted to: “NATO troops will be stationed in Kosovo to supervise the agreement.” The status-of-forces provisions granted NATO personnel very broad rights throughout the FRY. The British House of Commons Foreign Affairs Committee later examined this point in detail. It treated Appendix B as a serious intrusion into Yugoslav sovereignty and criticized the inclusion of such a broadly drafted status-of-forces regime.[9] Claims that: “Appendix B did not exist” or “NATO asked only for entry into Kosovo” are therefore incorrect. Does this prove that Rambouillet was designed to be rejected?

No. That is one interpretation. It is not an established fact. The British parliamentary inquiry heard arguments that such broad NATO rights would have been unacceptable to almost any sovereign state.[9] It also heard negotiators who said Appendix B was not itself the dominant issue during the talks and found no simple proof that this provision alone caused the negotiations to collapse.[9] Appendix B was real and unusually broad. Its existence does not automatically establish a concealed motive: “the agreement was drafted only so Milošević would reject it and NATO would have a pretext for war.” A claim about motive requires more evidence than the existence of a severe clause.

Why did Belgrade reject NATO troops in Kosovo?

For the Milošević government and the Yugoslav delegation, the issue was sovereignty. Kosovo was, under their legal position, part of Serbia and the FRY. NATO was a foreign military alliance of which Yugoslavia was not a member. Accepting NATO troops, command structures, jurisdictional immunities, and broad movement rights would therefore mean accepting foreign military power on Yugoslav territory. American diplomatic records themselves recognized that Belgrade rejected Rambouillet in significant part because it required a NATO-led military presence.[10] This was not an invented Serbian objection. It was a genuine central dispute. Belgrade did not unconditionally accept every political provision either.

The opposite myth also needs correction. It is often said: “Serbia accepted all political provisions and rejected only Appendix B.” Records from the negotiations show a more complex position. On 23 February Madeleine Albright said that the Serbian side had engaged with the political document and accepted many elements of Kosovo self-government, but that its acceptance was not unconditional.[11] When talks resumed in Paris in March, the U.S. side said the Serbian delegation reopened some issues that had previously been discussed politically.[12] The military implementation issue was the largest and clearest obstacle. But the dispute cannot responsibly be reduced to one paragraph of Appendix B alone.

18 March: the Kosovo-Albanian side signs, Belgrade does not

Negotiations resumed in Paris on 15 March. The Kosovo-Albanian delegation accepted and signed the proposed agreement. The Yugoslav/Serbian delegation did not.[12] From that moment the Western powers had a politically powerful image: one side had accepted their proposal; the other had not. Belgrade’s interpretation was different: the text before it was not merely a constitutional arrangement for Kosovo, but also a military regime of NATO deployment and broad privileges that it considered unacceptable. Both facts belong in the historical record.

March: the OSCE mission leaves Kosovo. The Kosovo Verification Mission withdrew on 19 March. The stated reason was that the security situation had deteriorated to a point where unarmed international observers could no longer be kept safely in the field.[10] The mission had been incomplete and often unable to stop violence. But it also represented more than a thousand international observers on the ground.[13] When it withdrew, the principal direct international monitoring mechanism disappeared just as forces were concentrating for renewed operations.

20–23 March: the final stage before the bombing. After the KVM withdrawal, Serbian and Yugoslav operations intensified sharply. U.S. diplomatic records reported large concentrations of forces in and around Kosovo and renewed displacement operations.[10] Richard Holbrooke made one final trip to Belgrade. He did not persuade Milošević to accept the terms. On 23 March, Javier Solana instructed Supreme Allied Commander Europe Wesley Clark to initiate air operations. On 24 March 1999, the attacks began.[14]

Did the UN Security Council authorize the bombing?

No. This is one of the most important legal facts in the story. Before the war the Security Council had adopted Resolutions 1160, 1199, and 1203. They condemned violence, demanded an end to fighting, called for a political settlement, endorsed verification arrangements, and required implementation.[1][2] None contained an explicit authorization stating that member states could use military force against the FRY. NATO therefore began the air campaign on 24 March without the ordinary form of explicit Security Council authorization for the use of force. That does not mean the Security Council later “authorized” it either.

On 26 March, Russia, Belarus, and India proposed a resolution demanding an immediate halt to NATO’s use of force. The draft failed by 3 votes in favor to 12 against, with no abstentions.[15] NATO supporters cited the result as evidence that most Security Council members did not want to halt the intervention. Legally, however, the distinction matters. Rejecting a resolution that would condemn or stop military action is not the same as adopting a resolution that authorizes that military action. The Security Council never explicitly authorized the air campaign. Humanitarian intervention: political justification and legal objection.

NATO justified its decision by arguing that intervention was necessary to avert a humanitarian catastrophe and compel Belgrade toward a political settlement.[14] The United States, United Kingdom, and other members argued that repeated resolutions, agreements, and threats had failed. Russia, China, and other states argued that the use of force without Security Council approval violated the UN Charter and the sovereignty of the FRY.[15] This is not a dispute that can be resolved by a slogan. The legal controversy was real. The humanitarian rationale was a genuine political argument advanced by NATO states. The explicit Security Council authorization was absent. June 1999 shows a major legal difference.

When the war ended in June, the Security Council adopted Resolution 1244. It was adopted under Chapter VII. It explicitly authorized international civil and security presences in Kosovo. It provided for: the withdrawal of Yugoslav and Serbian military, police, and paramilitary forces; an international security presence with substantial NATO participation; an interim UN administration; demilitarization of the KLA; and a process of substantial self-government in Kosovo.[16] It also reaffirmed the sovereignty and territorial integrity of the FRY.[16] The legal framework in June was therefore different from that of 24 March. In June there was an explicit Security Council mandate for the international presence.

In March there was no equivalent authorization for the bombing campaign. The postwar settlement was not simply Appendix B implemented. The June settlement created a very powerful international military presence in Kosovo, under a UN mandate and with major NATO participation.[16] It did not simply reproduce Appendix B’s FRY-wide rights of unrestricted NATO movement. That does not prove Appendix B was a deliberate trap. It does show that the arrangement that actually ended the war differed from the Rambouillet proposal in the breadth of NATO rights outside Kosovo. Appendix B is therefore a legitimate and important historical question, not a marginal footnote.

What can the evidence support about Račak?

The KLA was present in Račak. The village was not militarily neutral. The operation followed prior KLA attacks. Serbian police officers had been killed days earlier.[4] MUP and the Yugoslav Army jointly participated. This is a finding of the Đorđević judgment.[4] At least 45 Kosovo Albanians were killed. The judicial reconstruction also found numerous close-range wounds.[4]

The Finnish forensic investigation did not support the theory of staged fighter bodies. It described at least 40 unarmed civilians and did not find evidence of systematic postmortem manipulation.[5] ICTY proceedings did not all treat Račak identically. The Milutinović trial did not litigate the specific charge; the later Đorđević judgment did reconstruct the event in detail.[4][6] Walker’s political condemnation preceded the forensic process. That is a legitimate criticism of procedure.

That does not mean later evidence disproved civilian killings. It did not. What can the evidence support about Rambouillet? It was a real proposal for extremely broad Kosovo autonomy. It was not an empty diplomatic document. It did not provide immediate independence. Kosovo would formally have remained within the FRY during the interim period.[8] Final status after three years was deliberately ambiguous.

The text referred to the will of the people and an international meeting, not automatic independence.[8][9] The United States separately signaled support for a future referendum interpretation. That strengthened Serbian fears that autonomy could become a transitional stage toward secession.[9] The Kosovo-Albanian side did not immediately accept the proposal in February. KLA demilitarization was a major internal obstacle.[9] It signed in March. The Yugoslav/Serbian delegation did not.[12] Appendix B genuinely granted NATO very broad FRY-wide rights.

This is not a myth.[8] Its existence does not prove that Western negotiators deliberately designed the agreement to be rejected. The evidence permits debate but not a definitive motive claim.[9] What can the evidence support about the beginning of NATO’s bombing? Diplomacy occurred under an explicit NATO threat of force. Authority for possible air strikes existed before Rambouillet finally failed.[3] Belgrade did not accept the agreement. That is a fact.

The agreement also included politically difficult military terms. Especially NATO deployment and Appendix B. NATO began bombing on 24 March without explicit Security Council authorization. That is a fact.[15] Two days later the Security Council rejected a draft demanding an immediate end to NATO attacks. That is also a fact.[15] The rejection of that draft was not the same as legal authorization for the attacks. They are different legal acts.

The greatest error is to select only one document

If we show only Račak, we get the story of humanitarian intervention. If we show only Appendix B, we get the story of a pre-designed ultimatum. If we show only KLA attacks, we get the story of a state fighting terrorism. If we show only judicial findings of state crimes, we get a story without an Albanian armed actor. The real political history is more uncomfortable. The KLA used violence. Serbia and the FRY conducted major and often disproportionate security operations. Račak involved real civilian deaths and became a political catalyst.

Rambouillet offered extensive autonomy. Its military provisions also contained clauses that strongly affected FRY sovereignty. Belgrade did not accept the package. The Kosovo-Albanian side accepted it after internal disputes. NATO then chose force without explicit Security Council authorization. All of those propositions can be true at the same time.

Rambouillet is the boundary between diplomacy and war. From here the story moves to what happened after 24 March. What did NATO bomb? What was the military objective? How many civilians died? RTS. Grdelica. Niš. Varvarin. The Chinese Embassy. Cluster munitions. Bridges. The electricity grid. Industrial facilities. And at the same time: what happened to Kosovo Albanians on the ground; how many were expelled; what courts later found about coordinated deportation campaigns; and whether the bombing halted violence or initially accelerated it. 78 Days: NATO’s 1999 Bombing of Yugoslavia — Military Targets, Civilian Casualties, and the War in Kosovo