To date, most of the research on sexual violence (SV) has focused on assaults perpetrated by men against women. However, the prevalence of female perpetration of SV also warrants attention. Although the number of assaults perpetrated by women has been increasing, there is a lack of research on variables that may influence female-perpetrated SV. The confluence model of sexual aggression is a well-established framework to predict the risk of male SV perpetration, which includes the core constructs of hostile masculinity and impersonal sex. To investigate risk factors for female SV perpetration, we conducted a cross-sectional study to identify the most salient potential predictors of SV, including the primary constructs of the confluence model, in a sample of heterosexual female college students. A series of questionnaires about sexual experiences and gender roles were completed by 191 female undergraduates between the ages of 18 and 36 at a university in the northeastern United States. The measures were modified to reflect female pronouns and terms, and the novel predictor of hostile femininity was included, as an analog to the hostile masculinity construct in the confluence model. A series of random forest analyses indicated that both hostile femininity and impersonal sex were potential predictors of female SV perpetration. We also found sex role stereotyping, own sex role satisfaction, and acceptance of interpersonal violence to be among the strongest potential predictors. While these findings suggest that the confluence model may be applicable to female SV perpetration, there are additional factors that need to be considered to more fully understand the predictors of female-perpetrated SV.