The primary objectives of this article are to answer these two questions. A review of research findings indicates that men are more likely to beat women they live with than those they have married. The theoretical accounts of Goode and Gelles are used as a starting point for an explanation of this finding. Implicated in this explanation are social factors operating at the level of society (group membership), a disjunction between patriarchal societal norms and their existential basis in households (employment patterns), relational norms (ambiguity and flexibility), and social and economic costs. These variables influence woman abuse differentials by either increasing the amount of strain/stress or decreasing the likelihood of effective social regulation. 2 tables, 5 figures, appendix, 101 references. (Author abstract)