At Time 1, data were collected from a convenience sample of 190 college women. A total of 134 women were recruited from sororities (70 percent), and 56 (30 percent) were recruited from general psychology courses. At Time 2, data were obtained from 126 returning participants (66-percent completion rate). At both assessments, participants completed self-report questionnaires that assessed their history of interpersonal victimization, cognitions regarding risky activities, expected or actual involvement in a variety of risk-taking behaviors, and alcohol use. Questions pertinent to a participant's history of sexual victimization were behaviorally specific. Responses to questionnaires at Time 2 were matched to Time 1 responses. A total of 50 (26 percent) participants reported a history of some type of sexual victimization at Time 1 and were placed in the victim group. Of the 126 returning participants, 30 (24 percent) reported a history of some type of sexual victimization at Time 1 and were placed in the victim group. Sixteen new victimizations (12.7 percent) were reported to have occurred during the 5 1/2 months between assessments. Fifteen women reported an attempted rape, and 1 woman reported a completed rape. Of the 16 newly victimized women, 5 had reported a history of sexual victimization at Time 1. Discriminant function analysis indicated that alcohol use and expected involvement in risky activities at Time 1 were associated with new sexual victimization at Time 2. Hierarchical regression analysis determined that these two factors present at Time 1 predicted the frequency of involvement in risky sexual activities at Time 2. The implications of these findings for future research are discussed. 4 tables and 34 references