The study used existing data originally collected by Boyer and Fine (1992) in their longitudinal study of pregnant and parenting adolescents. Earlier uses of the data focused on the link between sexual victimization and adolescent pregnancy. Because of the current study's interest in predicting victimization and revictimization, the sample for analysis was limited to the 315 mothers who completed the survey at both time points. Analysis examined whether there were significant differences between those who completed the second survey and those who did not, based on information from the first survey. The survey contained items to describe subjects' demographic characteristics, family background, pregnancies, and parenting histories. Detailed sexual histories and behavioral indicators of sexual victimization, childhood maltreatment, and neglect were collected. Additional standardized scales on parenting, self- concept, and sexual attitudes were part of the survey. The dependent variable of interest was defined as any sexual victimization reported at Year 2 that occurred since the first survey. Using multivariate logistic regression analysis, data from Time 1 were used to predict victimization reported in the year between Time 1 and Time 2. Three factors -- any sexual victimization reported at Time 1, previous victimization that involved rape using force, and ever having been hit by a partner -- increased the likelihood of reported sexual victimization at Time 2. Three other factors were found to reduce the risk of victimization: having been victimized more than 1 year prior to the first survey; being pregnant at Time 1; and reporting satisfaction with social relationships. Additional analyses were used to investigate risk factors for revictimization. 5 tables, 24 references, and appended description of victimization variables