COMPARED ON A 31-VARIABLE TEST BATTERY AND THE PERSONALITY INVENTORY, THE 40 VIOLENT AND 40 NONVIOLENT PRISONERS DIFFERED SIGNIFICANTLY IN THEIR RESPONSES TO BOTH TESTS. SUBJECTS COULD BE CLASSIFIED CORRECTLY AS VIOLENT OR NONVIOLENT WITH 95-PERCENT ACCURACY USING THE NEUROPSYCHOLOGICAL TEST BATTERY ALONE. THE PERSONALITY INVENTORY ALONE CORRECTLY CLASSIFIED 79 PERCENT. THE DIFFERENCES BETWEEN THE TWO GROUPS INVOLVE COGNITIVE, LANGUAGE, PERCEPTUAL, AND PSYCHOMOTOR ABILITIES. IN EACH INSTANCE, THE NONVIOLENT GROUP PERFORMED BETTER THAN THE VIOLENT GROUP, WHICH SUGGESTS THAT THE VIOLENT SUBJECTS HAVE IMPAIRED BRAIN FUNCTION, AT LEAST INSOFAR AS THIS MAY BE INDICATED BY THE TESTS USED. WHILE NOT CONFIRMING BRAIN DISFUNCTION IN VIOLENT OFFENDERS, THE RESULTS DO INDICATE THAT THE NERVOUS SYSTEM INTEGRATION AND CONTROL NECESSARY TO PERFORM ACCURATELY ON THESE TESTS ARE WEAK RELATIVE TO THE NONVIOLENT GROUP. THIS IS CONSISTENT WITH EARLIER REPORTS THAT LINK BRAIN DISFUNCTION TO VIOLENCE. BY SHOWING THAT THE LACK OF CONTROL PRESENT IN THE LIVES OF VIOLENT PERSONS IS NOT SIMPLY A FUNCTION OF GROSS PERSONALITY TYPE, BUT ALSO MAY BE OBSERVED IN MORE MOLECULAR ASPECTS OF COGNITION, MEMORY, AND MOTOR BEHAVIOR, THESE DATA SUGGEST AN ORGANIC CONTRIBUTION TO THE POOR CONTROL SEEN IN SUCH PERSONS. TABULAR DATA AND REFERENCES ARE INCLUDED. (KBL)