Objective: This study presents an evaluation of a novel procedure designed to measure bystander behavior to prevent sexual violence. Method: Ninety-one college students participated in simulations conducted in an immersive virtual environment. The simulations were designed to elicit and allow measurement of bystander behavior in situations that could presumably escalate to sexual violence. Participants also completed self-report measures of responsibility for intervening, efficacy for intervening, and intent to intervene. In addition, self-reported bystander behavior was assessed approximately 2 months later. Results: Bystander behavior in the simulations was positively correlated with (a) responsibility for intervening, (b) efficacy for intervening, (c) intent to intervene, and (d) self-reported bystander behavior. Conclusion: Our findings support the validity of this virtual simulation procedure and lay the groundwork for future research designed to evaluate effects of bystander intervention programs and to study determinants of bystander behavior. (PsycINFO Database Record (c) 2016 APA, all rights reserved)