This study explores gender differences in memories for graphic media violence. More specifically, this project measures the level of complexity in men’s and women’s situation models for violent movies and TV programs seen in the past. Using the theory of vivid media violence, negative emotions, accessibility, and memory vividness are tested as predictors of situation model complexity. In a nationwide survey, 254 participants wrote essays describing memorable violent movies and TV programs. The essays suggest more similarities than differences among men’s and women’s situation models, although men described more details about blood and gore. Among both men and women, negative emotions at the time of exposure indirectly predicted situation model complexity as mediated by accessibility and memory vividness. Implications for scholars studying media violence effects are discussed.