To determine the association between fatal pulmonary embolism and use of antipsychotic drugs in a Swedish medicolegal autopsy series. Persons aged 18-65 years and subjected to a medicolegal autopsy in 1992-2005 were selected. On the basis of external cause of death, determined by the forensic pathologist, unnatural deaths (including fatal intoxications) were excluded and participants in whom pulmonary embolism was the cause of death were identified. Use of antipsychotics was based on the results of the postmortem analyses and categorized as use of high-potency first-generation antipsychotics, low-potency first-generation antipsychotics, second-generation antipsychotics or no use of antipsychotics. Logistic regression analyses were performed. Use of antipsychotics was verified in 538 of the 14,439 included participants. Pulmonary embolism was recorded as the cause of death in 279 participants and 33 of these used antipsychotics. Use of low-potency first-generation antipsychotics and second-generation antipsychotics was significantly associated with fatal pulmonary embolism (adjusted odds ratio: 2.39, 95% confidence interval: 1.46-3.92 and 6.91, 95% confidence interval: 3.95-12.10, respectively). Out of 26 participants classified as high-potency first-generation antipsychotic drug users, none had pulmonary embolism as the cause of death. Pulmonary embolism was overrepresented among medicolegal autopsy cases identified as users of low-potency first-generation and second-generation antipsychotics.