The sample included 100 men and women charged with murder or nonnegligent manslaughter who were referred for psychiatric evaluation between January 1980 and December 1988. More than half of the subjects actively abused drugs at the time of their crimes. Each subject was interviewed and examined, and interviews were also conducted with family members, friends, employers, teachers, and therapists. Active substance abuse was defined as use on three or more occasions during the 30-day period preceding the homicidal event (five or more occasions for alcohol). Demographic and other discriminating factors were employed to examine the hypothesis that murderers did not constitute a homogeneous population and that subgroups differed in their abuse patterns. Cluster analysis techniques were applied to the study population. Using a set of 13 proximate causal factors, a typology of seven distinct homicide profiles was created. Two of the seven profiles exhibited extremely high substance abuse and intoxication rates, three others intermediate rates, and two profiles very low rates. Moreover, different substances were prime offenders in different profiles. Alcohol was the most common substance of abuse in homicide cases. Findings revealed that substance abuse is an important etiological contributor to some murderer types.