Results indicated that violent victimization was significantly higher for the incarcerated group; 47 percent experienced violent victimization within 6 months of incarceration compared to the 1.4 percent of women in the NCVS community sample who experienced violent victimization. Findings for the NCVS community sample fit nicely within the prediction of routine-activities theory, however this theory was not as well suited to predicting the violent victimization of women living in marginalized circumstances. Unlike the experiences of the NCVS community group (mainstream women), partnering with a man does not protect marginalized women from violent victimization, quite the opposite as a matter of fact. Moreover, the findings indicated that marginalized women might be better able to negotiate certain types of risk than mainstream women, such as living in poverty and moving often. These findings challenge the predictions of routine-activities theory. Research involved comparing risk and protective factors for violent victimization among female respondents to the National Crime Victimization Survey (NCVS) with those of female offenders incarcerated in Baltimore, MD. Participants were 351 female inmates of the Baltimore City Detention Center who were interviewed using a computerized life-event calendar to collect monthly information about women’s life circumstances and experiences with violence in the 3 years prior to incarceration. The NCVS sample included 646,549 females who were interviewed with the redesigned NCVS between January 1992 and June 2000 about their characteristics and experiences as crime victims. Logistic regression models were used to analyze the data. Future research should continue to explore violent victimization among marginalized populations of women in order to inform theories of violence against women. Table, figures, appendix, notes