Data for the study were collected during the period from March 1987 to January 1992 from psychiatric evaluations of 63 defendants performed at the request of the Los Angeles County Superior Court. Of these 63 persons, 25 suffered from psychosis and formed the study sample. Information was obtained by record review of each subject. The 25 defendants allegedly killed 27 persons. The group contained 3 females (12 percent) and 22 (88 percent) males. Diagnostically, there was a clear majority (60 percent) of the psychotic sample who suffered from schizophrenia. This is in marked contrast to the prevalence of schizophrenia and bipolar disorder found among the community-at-large in which the ratio of schizophrenia to bipolar disorder would be no higher than 2 to 1. Schizophrenia, with its preponderance of psychotic symptoms, apparently exerts greater influence than bipolar disorder with its mood symptoms in the genesis of homicidal violence. Of the 27 victims, the murder defendant knew his/her victim in a majority (51.9 percent) of the cases, consistent with the national figure of all reported U.S. homicides. Also, 44.4 percent of the victims were killed in either their own home, the home they shared with the defendant, or the defendant's home. This finding suggests that in approximately one-half of the murders by psychotic persons, familiarity plays an important role in who is murdered and where the murder occurs. On the other hand, the sample also reported murdered persons unfamiliar to them in a substantial portion (44.4 percent) of the cases. 21 references