With a rise in mass casualty incidents, training in hemorrhage control using tourniquets has been championed as a basic-and lifesaving-procedure for bystanders and medical professionals alike. The current standard for training is in-person (IP) courses, which can be limited based on instructor availability. Virtual reality (VR) has demonstrated the potential to improve the accuracy of certain medical tasks but has not yet been developed for hemorrhage control. The objective of this study was to evaluate the efficacy of a VR hemorrhage trainer in learner retention of tourniquet application when compared to traditional IP instructor teaching among a cohort of emergency medicine residents practicing in a Level I trauma center. This was a prospective, observational study of 53 emergency medicine residents at an inner-city program. Participants were randomly assigned to either the control or the VR group. On Day 0, all residents underwent a training session (IP vs. VR) for the proper, stepwise application of a tourniquet, as defined by the American College of Trauma Surgeons. Each participant was then assessed on the application of a tourniquet by a blinded instructor using the National Registry Hemorrhage Control Skills Lab rubric. After 3 months, each resident was reevaluated on the same rubric, with subsequent data analysis on successful tourniquet placement (measured as under 90 s) and time to completion. Of the 53 participants, the IP training group had an initial pass rate of 97% (28/29) compared to 92% (22/24) in the VR group (p = 0.58). On retention testing, the IP training group had a pass rate of 95% (20/21) compared to 90% (18/20) in the VR group (p = 0.62). Stratifying the success of tourniquet placement by level of resident training did not demonstrate any statistically significant differences. In this pilot study of emergency medicine residents, we found no significant differences in successful hemorrhage control by tourniquet placement between those trained with VR compared to a traditional IP course among emergency medicine residents. While more studies with greater power are needed, the results suggest that VR may be a useful adjunct to traditional IP medical training.