BACKGROUND: In Sweden 20 000 cases of assault against women are reported to the police every year. METHOD: All data on the perpetrators of spousal homicide in Sweden between 1990 and 1999 were investigated (n = 164). A control group of all other perpetrators of homicide in Sweden during the same period, i.e. cases of homicide not committed in the context of spouse violence (n = 690) was used. All verdicts, as well as all material in the police investigations, including interviews with all of the police investigators, were analysed. Copies of police examinations of the suspects, and forensic reports from the autopsies, were also examined. Data on all registered criminality were collected from the National Police Register, and in cases where the perpetrators had been subject to forensic psychiatric examinations, those reports were obtained from the Swedish National Board of Forensic Medicine. In addition, the Psychopathy Checklist: Screening Version scores were rated from the forensic psychiatric examinations. RESULTS: There was a four times higher suicide rate among the spousal homicide perpetrators (24%, n = 40) compared with the perpetrators in the control-group (6%, n = 39, chi-squared = 55,42 df = 1, p < 0.001). Consequently, suicidal ideation must be considered as an important risk factor for spousal homicide. In 79% of the cases the spousal homicide perpetrators were subject to forensic psychiatric examinations. All except 5% were diagnosed with at least one psychiatric diagnosis, and 34% were sentenced to forensic psychiatric treatment. If it is assumed that the psychiatric morbidity was high in the 24% of the perpetrators who committed suicide, then 80% of all perpetrators of spouse homicide during the study period can be characterized as mentally disordered. 'Psychopathic' perpetrators, who generally are over-represented in most violent criminality, were comparatively uncommon. Only seven (4%) in the study group met the diagnostic criteria for psychopathy as measured with the PCL:SV. DISCUSSION: The group of spouse killers studied here fits the dysphoric/borderline group of spouse assaulters. This is a group that may benefit from treatment. Perhaps police officers could help identify this kind of spouse assaulter before a fatality occurs.