The findings from the study suggest a number of conclusions about the links between structural gender inequality and the likelihood of getting out of violent relationships and about the ways in which our understanding of domestic violence is enhanced by examining different types of violence. First, the findings suggest that men and women who experience domestic violence are unequally situated in their marriages in terms of economic dependency on a partner which affects their likelihood of marital dissolution. The economic dependency suppresses women’s ability to get out of relationships in which they are victimized. The hypothesis that sex differences in the likelihood of dissolving a violent relationship would be mediated by depression was not supported. The number of young children in the household was positively linked to marital dissolution among women but negatively linked to marital dissolution among men. This finding suggests that men who are both perpetrators and victims of severe violence were most reluctant to initiate a divorce or separation when they had young children. However, the data cannot explain why this pattern occurs. Divorce literature has identified violence as a predictor of relationship dissolution, but this literature has not examined the ways this connection may differ for women and men. This study investigated the question of who leaves a violent relationship using a nationally representative sample of married couples that includes information about partner violence victimization collected from both men and women. Table, notes, references