A statistical evaluation of 150 homicidal patients was presented. The ratio of male: female patients was 4:1. One hundred and fifty-seven persons were killed by 150 patients. Of the victims 73 (47%) involved family members. Thirty-two male patients (27%) killed their wives, whereas 6 wives (19%) killed their husbands. Fifteen men (13%) killed their girlfriends; 7 female patients (22%) killed their boyfriends. fanticide and murder of children amounted to 22 (14%) of the victims (40% of the female murderers killed their infants and/or children).The cases may, grosso modo, be classified as follows(13):1. Those in which the homicide was the direct offspring of delusions (paranoid group as described in this paper): 40%;2. Those in which the homicide was committed during a paroxysm of insanity and/or was committed by manifestly insane persons, from motives and conditions which might influence the "sane mind" such as anger, revenge, jealousy, etc. : 32.6%;3. Those in which the insanity was said to have developed after the homicide which was committed for motives such as robbery (9), morbid jealousy, sex crimes, anger, "cuckolding-reaction," revenge, etc. (psychosis w/psyohopathic personality): 27.3%.In most of the cases classified under Group 1 (40%) forebodings of the malignant nature of patients' mental illness were present over a long period; the malignant nature of persistent delusions of infidelity in the involutional psychosis has been previously described(11). The remaining diagnostic groups (described under 2) usually unfold the symptoms of their mental illness over a long period. Prevention in Group 3 is often more a penal than a psychiatric problem.Intemperate use of alcohol was a contributing factor in 34% of our cases, whereas severe intoxication at the time of the crime could be elicited from the records in 12.5% of the cases.It is suggested that an interdisciplinary team consisting of specialists collaborating with each other in cultural, constitutional, psychological, neurological, genetic and dynamic studies run their data concerning homicidal patients through a computing machine and establish prediction tables for the various mental illnesses which would be used as a yardstick for prevention. I would like to conclude with A. Wikler's(12) definition and goals of psychiatric research: that sort of activity which is directed to the accumulation of data that are useful for the purpose of prediction and control of observable phenomena. The phenomena with which the psychiatrist is concerned are largely those related to the interpersonal activities of human beings. He wishes to be able to predict such behavior and where necessary for the patient, society, or both, to alter it.