Two theories of rape reporting, the Classic Rape perspective and Black's Theory of the Behavior of Law, are tested in this article. We offer the first comprehensive multivariate test of Classic Rape predictions among a nationally representative sample of victims, as well as the first test of Black's predictions for rape reporting. Through the construction of multinomial regression models, we are able to examine reporting patterns for both victims and third parties. Weapon use and physical injury consistently predicted reporting. The likelihood of victim reporting significantly increased when assaults occurred either in public or through a "home blitz," whereas place of assault did not affect the likelihood of third-party reporting. On the other hand, victim-offender relationship significantly affected the likelihood of third-party reporting but was not significant in the victim-reporting models. There were mixed findings regarding Black's stratification and morphology predictions, and we found no significant effects for culture, organization, or social control. Overall, these findings lend greater support to the Classic Rape perspective than to Black's model.