The study did not find support for the hypothesis that brain dysfunction was responsible for violent behavior in psychiatric patients independent of substance abuse. Results indicate that psychiatric inpatients with a history of community violence did not significantly differ on tests of executive control, attention, and general intelligence compared to nonviolent patients with similar demographics and psychiatric diagnoses. Patients with a history of violence did not demonstrate significantly higher rates on neurodevelopmental variables, including left-handedness, developmental learning disabilities, and childhood hyperactivity, nor did they demonstrate higher rates of neurological abnormalities on computerized tomography (CT) scans of the head and electroencephalogram (EEG) patterns in comparison to nonviolent patients. The only significant association between a neurological variable and violence was a history of closed-head injury; however, logistic regression analysis revealed that the strong association between comorbid substance abuse and violence could explain this relationship. These patients were not in the State neurorehabilitation program, and as such the relatively milder nature of the head injury might help to explain the overwhelming influence of substance abuse in this study. Data were collected from 63 patients who had been diagnosed with a major mental illness and had been hospitalized for 3 to 4 months at the time of assessment. Tables and references