Results indicated that nearly two-thirds of the sample had experienced intimate partner violence at some point during their adult life, with much of this violence being episodic and limited over time. Only 4 percent of participants were consistently involved in a battering relationship. Slightly less than one-third of participants who reported the battering had stopped returned to their violent relationship. Among the 280 women with complete data, the rate of recurring violence was 64 percent for women who reported current violence and was 44 percent for women who reported prior (lifetime) violence. In terms of risk markers for violence, child sexual molestation was significantly related to adult intimate partner violence, as were inadequate emotional support from nonprofessionals during adulthood, poor self-esteem, having a partner with substance abuse problems, and having a partner with a poor work history. Contrary to the results of previous studies, ethnicity was not significantly related to whether a woman was abused during the past year. Data were drawn from the Worcester Family Research Project (WFRP), which used an unmatched case-control design to recruit homeless and housed female heads of households to complete a series of interviews at three points between 1992 and 1997. The current analysis relied on data from 220 homeless mothers from the first set of interviews, 356 homeless mothers from the second set of interviews, and 327 homeless mothers from the third set of interviews. Interviews focused on personal history, physical and sexual victimization, personal assessments of social support resources, self-esteem, parental bonding, partner characteristics, and help-seeking behaviors. Multivariate analyses were used to identify risk markers. Future research should consider the contextual factors influencing violence among extremely poor women with children, such as neighborhood policing practices and community support for battered women. Tables, references