Objective: The goal of the current study was to examine how diversity of men’s social networks might reduce their sexual violence, either directly or indirectly via their hostile attitudes toward women, when taking measures to account for perceptions of peer attitudes concerning sexual violence. Method: A sample of male college students completed a series of questionnaires concerning their sexual behavior, social networks, and individual and perceived peer attitudes toward women and sexual violence. Because the hypothesized sexual violence outcome was positively skewed, a series of count-based regression models were tested, then the mediational path model was fit using the most appropriate regression method. Results: Zero-inflated negative binomial regression fit the data best, meaning sexual violence was estimated as both men’s likelihood to perpetrate sexual violence and the sexual violence rate among likely perpetrators. Findings suggest social network diversity reduces the likelihood of perpetrating sexual violence indirectly through men’s hostility toward women. Furthermore, social network diversity directly reduces the rate of sexual violence among possible perpetrators. Conclusion: Findings suggest men who are active in more social groups are less likely to be sexually violent, and if they are, they tend to commit fewer acts of sexual violence. Furthermore, hostile attitudes toward women function as a mechanism linking peer-network diversity to men’s likelihood of perpetrating sexual violence. Although more research is needed to confirm the temporal order of these effects, these findings may inform bystander intervention programs and social norms campaigns intended to reduce the prevalence of campus sexual violence.