There were no differences among the groups in racial composition, religious preference, or religious devoutness. Male abusers were less likely to be employed, to be in intact relationships, and to be less well educated. They were more likely to have witnessed abuse or experienced abuse as children, although that observation is more characteristic of abusers with alcohol problems. Measures of personality and psychopathology generally supported the hypothesis that abusive males would show greater elevations on test scales reflecting personality disorder and dysphoria and less conformity than nonbatterers. Alcohol abuse was related to greater batterer-nonbatterer differences. Implications of these findings and methodological concerns for future research are discussed. 38 references.