The findings did not indicate that women who witnessed parental violence as a child were more inclined to be victimized by their husbands, despite previous findings to the contrary. Witnessing violence as a child may be confounded with other more discriminating risk markers. However, abusive husbands were more likely to have witnessed parental violence than nonabusive husbands. In this multivariate study, experience of parental violence did not discriminate between wife assault victims of any kind from nonassaulted women. Although a woman's low self-esteem has been associated with an increased likelihood of wife assault, the independent contribution of this factor does not discriminate between victims and nonvictims. The two measures of status incompatibility used in this study -- religion and education -- did not differentiate between assaultive and nonassaultive relationships. The results of this four-group analysis of variance found that socioeconomic status did not discriminate between physically assaulted and non-physically assaulted wives, but that lower socioeconomic women were more likely to be severely assaulted. The only factor to discriminate physically assaulted wives from non-physically assaulted wives is martial conflict; it also differentiated the non-assaulted groups from the verbally assaulted group. Further research should focus on the perpetrators of wife assault, the dynamics of the relationships, and the social environment of the relationship. 3 tables, 21 references. (Publisher abstract modified)