Participants were 26 wife-abusers and 18 nonviolent men in therapy. Early exposure to marital violence and child abuse was assessed through interviews that were coded both qualitatively and quantitatively. The results indicated that batterers were more likely than comparison subjects to have been abused as children, to have witnessed their father beating their mother, and to have been disciplined as children with corporal punishment. Fathers were no more likely than mothers to have abused the men. There were no differences between groups in reports of parental child abuse toward siblings of the men. Differences between groups in exposure to violence were studied to explain why men who grew up in violent homes were not abusive toward their spouses. Socioeconomic status and family dynamics were examined to explain the results. 1 note. 55 references.