A population-based cohort of 167 Swedish homicide offenders from 1970-1980 for which causes of death during the following minimum of 22 years were investigated were studied using survival analysis. 29 suicides occurred during the follow-up representing 30% of the total mortality. A high proportion (72%) of suicides occurred early, within 2 years after the homicide. The suicides continued to accumulate during the following decades after the homicide, & the cumulative suicide risk was 18.6%. A very high proportion of violent methods (86%) was found. (26 refs.)