Using a representative sample of 2,703 Canadian women living in either a stepfamily or a biological family, this investigation assesses the extent of women's elevated risk for violence in stepfamilies relative to biological families as well as explanations for this relationship. Canadian women living in stepfamilies are shown to be twice as likely as their counterparts in biological families to experience violence. Differences between the two groups are greatest on some of the most severe forms of violence, suggesting that women in stepfamilies are at particular risk for severe violence. Institutional incompleteness (number of children; depression; alcohol consumption), duration of relationship, evolutionary psychology (sexual possessiveness; sexual jealousy; female employment; education compatibility) and selection factors (previous marriage/common-law union; previous partner violence; marital status) are applied and tested. Results show partial support for each explanation and that no explanation alone accounts for the disproportionate risk of violence in stepfamilies. Rather, a combination of elements from all explanations is required to account for the higher odds of violence against women in stepfamilies.