On February 28, 1986, Swedish Prime Minister Olof Palme and his wife left a Stockholm cinema and walked home without security. A gunman shot Palme in the back at 11:21 PM on Sveavägen, one of the city's busiest streets. He died within minutes. The investigation would become Scandinavia's longest and most expensive criminal case — involving over 10,000 named suspects, more than 20,000 witness interviews, and approximately $100 million in costs. In June 2020, Swedish prosecutors announced they were closing the case and naming Christer Pettersson — who had died in 2004 — as the killer. The evidence was circumstantial. The motive was unexplained. Almost no one was convinced.
At 11:21 PM on February 28, 1986, Swedish Prime Minister Olof Palme and his wife Lisbet were walking home from the Grand cinema on Sveavägen, one of Stockholm's busiest streets. They had dismissed their security detail earlier that evening to watch a movie in peace. A man approached from behind and fired two shots at close range. The first bullet struck Palme in the back, severing his spinal cord and penetrating vital organs. The second grazed Lisbet's shoulder. Palme collapsed on the sidewalk. He was pronounced dead at Sabbatsberg Hospital forty-five minutes later.
Sweden had never experienced a political assassination in modern times. The country prided itself on being a peaceful social democracy where political leaders walked freely among citizens. Palme himself often refused security protection, believing it created unnecessary distance between government and people. That decision would prove fatal — and the investigation into his death would become one of the longest, most expensive, and most controversial criminal cases in Scandinavian history.
The Swedish police response was immediately compromised by procedural failures that would haunt the investigation for decades. The crime scene was not properly secured. Thousands of curious onlookers walked through the area before forensic teams could complete their examination. Potential evidence was contaminated or lost. Witnesses were not systematically interviewed, and conflicting accounts were never reconciled. The murder weapon — believed to be a .357 Magnum revolver based on ballistic analysis — was never found despite extensive searches.
Hans Holmér was appointed lead investigator within hours of the shooting. Within weeks, he became convinced that Kurdish separatist groups were responsible. His theory was based on intelligence suggesting that Palme's government had recently denied asylum to Kurdish activists and that members of the PKK (Kurdistan Workers' Party) were active in Sweden. Holmér authorized extensive surveillance of Kurdish communities, wiretaps on hundreds of phones, and raids on cultural centers. Several Kurdish nationals were arrested. None were charged.
The Kurdish theory consumed the first critical year of the investigation. Holmér's fixation meant that other leads were ignored or inadequately pursued during the period when evidence and witness memories were freshest. Swedish media later revealed that witnesses who reported seeing suspicious individuals near the cinema — individuals who did not match Kurdish physical descriptions — were not thoroughly interviewed because their accounts didn't fit Holmér's theory.
"We lost the first year chasing the wrong people while the real leads went cold. That mistake can never be corrected."
Jan Danielsson — Swedish Police Review Commission, 1987In March 1987, the Swedish Bar Association filed formal complaints about civil rights violations during Holmér's investigation. He was removed from the case. Years later, Holmér admitted the Kurdish theory was likely wrong but defended his actions by claiming he had followed what appeared to be credible intelligence assessments at the time.
In December 1988, Swedish police arrested Carl Gustav Christer Pettersson based on a tip from an acquaintance who claimed Pettersson had drunkenly suggested he was involved in the murder. Pettersson was a 41-year-old man with a substantial criminal record, including a 1970 manslaughter conviction for killing a man with a bayonet during a street fight. He had a history of drug abuse and alcoholism.
Police showed Lisbet Palme a lineup that included Pettersson. Ten days after her husband's murder — still in shock and grief — she identified Pettersson as resembling the shooter. This identification became the centerpiece of the prosecution's case. There was no physical evidence linking Pettersson to the crime scene. No witness had seen him near the cinema that night. The murder weapon had not been found. Prosecutors could not establish a clear motive explaining why a petty criminal with no political interests would assassinate the Prime Minister.
The trial was brief. Pettersson maintained his innocence. His defense attorneys argued that the widow's identification was unreliable given the traumatic circumstances, poor lighting conditions on the street, and the time elapsed between the murder and the lineup. The prosecution presented no forensic evidence, no ballistic matches, no witnesses placing Pettersson at the scene. The conviction rested on a single witness who had seen the killer's face for seconds during the most traumatic moment of her life.
In October 1990, the Svea Court of Appeal overturned the conviction. The appeals court ruled that Lisbet Palme's identification alone was insufficient to establish guilt beyond reasonable doubt. The judgment noted that witness identification under traumatic circumstances is inherently unreliable, especially when significant time has passed and when the witness has been shown photos of the suspect before the lineup. Pettersson was released and awarded 300,000 Swedish kronor (approximately $50,000) in compensation for wrongful imprisonment.
While the Pettersson trial dominated headlines, a different theory was emerging from an unexpected source. In the mid-1990s, as apartheid ended and South Africa began its Truth and Reconciliation Commission process, former intelligence operatives began testifying about covert operations conducted during the 1980s. Multiple witnesses described Olof Palme as a high-priority target.
The motive was clear and documented. Between 1982 and 1986, Palme's government provided substantial material support to the African National Congress, including financial aid, diplomatic protection, and facilitation of weapons transfers through Swedish territory. Sweden offered asylum to hundreds of ANC members and allowed the organization to maintain offices in Stockholm. From the perspective of South African military intelligence, Palme was actively supporting terrorism against the South African state.
In August 1996, former South African intelligence operative testimony before the Truth and Reconciliation Commission confirmed that planning documents existed for operations against Palme. Craig Williamson, a major in South African intelligence who had run covert operations across Europe, testified that Palme's assassination was discussed at high levels within the security establishment. He could not confirm whether the operation was ultimately executed by South African operatives or whether someone else killed Palme first.
The theory gained further credibility when documents declassified in 2007 revealed that South African military intelligence had compiled extensive surveillance files on Palme's movements, his meetings with ANC leadership, and Swedish weapons shipments to southern Africa. The files showed active monitoring of Palme dating back to at least 1983.
Eugene de Boisanger, a former French Foreign Legion officer working for South African intelligence, was specifically named by multiple sources as having been tasked with organizing the assassination. Swedish journalist Lars Borgnäs presented evidence in 1996 suggesting that de Boisanger was in Scandinavia in February 1986 and had met with operatives in Copenhagen shortly before the murder. De Boisanger denied involvement before his death in 1992. Swedish authorities investigated the South African connection extensively but never filed charges, citing insufficient evidence to establish that South African operatives actually pulled the trigger.
Multiple witnesses reported seeing a man near the Grand cinema entrance in the minutes before Palme and his wife emerged. The descriptions were consistent: a man in his 40s, approximately 180 centimeters tall, wearing a dark coat, appearing to watch the theater entrance. This individual was seen leaving the area quickly after the Palmes exited. Police created a composite sketch based on these descriptions. The man depicted did not resemble Christer Pettersson.
This unidentified individual became known as the "Skandia Man" after the insurance company whose building stood near the cinema. Despite extensive publicity and appeals for information, the Skandia Man was never identified or interviewed. Some investigators believe he was a lookout for a coordinated operation. Others suggest he may have been an innocent bystander who left to avoid involvement. A third theory holds that he was the actual shooter and that Lisbet Palme's identification of Pettersson was simply wrong.
The failure to identify the Skandia Man represents one of the most significant gaps in the investigation. When prosecutors announced in 2020 that they were closing the case, they did not address the Skandia Man sightings or explain why witness testimony describing someone other than Pettersson was considered irrelevant.
On June 10, 2020, Chief Prosecutor Krister Petersson held a press conference in Stockholm. After three years of reviewing the case, he announced that the investigation was being closed. He named Christer Pettersson — who had died in 2004 after a fall — as the person responsible for the assassination. The decision meant no trial would be held, no new evidence would be tested in court, and no definitive proof would be presented to the public.
Petersson's announcement was based on circumstantial evidence. He cited mobile phone tower data that allegedly placed Pettersson near the crime scene on the night of the murder — data that had not been presented at the original 1989 trial. He referenced witness testimony from that trial, primarily Lisbet Palme's identification. He could provide no clear motive. When asked why Pettersson would have targeted Palme, Petersson suggested it might have been a random act or a case of mistaken identity, though he offered no evidence supporting either theory.
The reaction was immediate and overwhelmingly negative. Palme's son Mårten stated that the family was not satisfied with the conclusion. Criminologists criticized the decision to close the case without addressing alternative theories. Journalists who had investigated the South African connection argued that evidence of a political assassination had been dismissed without adequate explanation. Legal experts questioned whether closing the case simply because the named suspect was dead met the standard for justice.
"To close this case without explaining the Skandia Man, without addressing the South African evidence, without establishing motive — this is not justice. This is bureaucratic convenience."
Jan Bondeson — Criminologist, Lund University, 2020Swedish law requires prosecutors to close cases if no prosecution is possible. Since Pettersson was dead, no trial could be held. Petersson defended his decision by stating that continuing the investigation indefinitely served no purpose if the available evidence could not support charges against a living person. Critics responded that if the evidence against Pettersson was insufficient to justify charges against anyone else, the case should remain open rather than closed with a declaration of guilt that could never be tested.
After 34 years and an estimated 1 billion Swedish kronor ($100 million) spent on investigation, certain facts are established. Olof Palme was shot on February 28, 1986, by someone who approached from behind on a public street. The killer escaped. The murder weapon was never found. No physical evidence definitively identified any suspect.
The Kurdish theory was thoroughly investigated and found to be baseless. The PKK had no demonstrated motive, and extensive surveillance of Kurdish communities produced no evidence of involvement. That theory is considered definitively disproven.
The Christer Pettersson theory rests on a single witness identification made under traumatic circumstances. A court initially accepted that identification as sufficient for conviction. An appeals court found it insufficient. No additional evidence has emerged in the three decades since his acquittal. The 2020 decision to name him as the killer was not based on new physical evidence but on a prosecutorial determination to close the case.
The South African theory has substantial circumstantial support. Documentary evidence confirms that South African intelligence targeted Palme, maintained surveillance files on his activities, and discussed assassination operations. Witness testimony from former operatives indicates that planning occurred. However, no evidence directly connects any South African operative to Stockholm on February 28, 1986, or establishes that the assassination was executed by South African agents rather than by someone else.
Alternative theories involving Swedish right-wing extremists, police officers with grievances against Palme's policies, or random violence have been investigated but remain speculative without supporting evidence.
The Palme assassination investigation failed at multiple structural levels. The initial crime scene management was inadequate. The lead investigator's fixation on an incorrect theory consumed the first year when evidence was freshest. The focus on Christer Pettersson in 1988-89 appears to have been driven by pressure to produce an arrest rather than by compelling evidence. The failure to identify and interview the Skandia Man represents a significant investigative gap. The reluctance to pursue international leads — particularly the South African connection — may reflect jurisdictional constraints or political considerations rather than evidentiary assessment.
The 2020 decision to close the case did not resolve these failures. It declared a conclusion without establishing proof that would satisfy judicial standards. It left alternative theories unaddressed. It provided no explanation for contradictory witness evidence. Most fundamentally, it substituted prosecutorial assertion for the judicial process that would have tested the evidence in an adversarial proceeding.
Sweden's tradition of open governance and social trust was shaken by Palme's murder. The failure to solve the case — or to solve it convincingly — represents an ongoing institutional failure. The investigation's mishandling, the wrongful conviction and later acquittal of Pettersson, the dismissal of credible alternative theories, and the ultimate decision to close the case by naming a dead man without trial have undermined public confidence in Swedish investigative capacity.
We do not know who killed Olof Palme. We have a prosecutorial assertion that cannot be tested, circumstantial theories that cannot be proven, and documentary evidence of planning by foreign intelligence services that may or may not have been executed. We have witness testimony describing individuals who were never identified. We have a murder weapon that was never found. We have investigative failures that are thoroughly documented but that do not lead to definitive conclusions about who benefited from those failures or whether they were merely incompetence rather than something more deliberate.
The case was closed in 2020 not because it was solved but because the legal system determined that continued investigation served no practical purpose. That determination may reflect institutional exhaustion, bureaucratic necessity, or genuine belief that no additional evidence would emerge. It does not reflect factual resolution.
Thirty-eight years after a gunman shot Sweden's Prime Minister on a Stockholm street, the question remains: who pulled the trigger, and why? The official answer — Christer Pettersson, motive unknown — satisfies almost no one who has examined the evidence carefully. The investigation that was supposed to restore confidence in Swedish institutions instead demonstrated their limitations. The case that was closed in 2020 remains, in every meaningful sense, unsolved.