Peer bullying is a widespread and longstanding problem in school settings. Teachers, students, administrators, government, and researchers alike have all tried to combat bullying through bullying prevention campaigns. One strategy used in bullying prevention campaigns is to call on bystanders in bullying situations to take responsible action. While many different forms of campaigns, including print and media campaigns, have aimed at trying to reduce the presence of bullying in schools by informing bystanders, there is still a need to find new strategies for reducing bullying behavior. One potential media form that could be used in bullying prevention campaigns is the use of virtual reality. Virtual reality simulations allow for a more immersive environment than other media forms, as the medium is capable of creating feelings of presence, various perspectives, and empathy in its users. This thesis reports results from a one-factor, three-condition laboratory experiment comparing responses to portrayals of a bullying situation in which users (N = 78) were placed in the perspective of a bystander in a bullying scenario across three different media stimulus conditions: a customized virtual reality condition, a non-customized virtual reality condition, and a video condition. The study compared effects of the media stimulus conditions on empathy, attitudes toward bullying victims and bullying, and anticipated future bystander behaviors, as well as presence and other outcomes related to perceptions of bullying. While it was hypothesized that the study would find stronger effects on empathy and anti-bullying bystander attitudes and anticipated behaviors among the VR conditions, and in particular the customized VR condition, the study found no differences between media stimulus conditions for any outcomes except a significant effect on empathy, with participants in the video condition tending to report more feelings of empathy for participants than participants in the other two conditions, and perceptions of bullying as a problem in the participants' school, again with scores highest in the video condition. This pattern of results was further explored in a follow-up qualitative focus group study (N = 10), in which trends from two focus group sessions featuring 10 participants indicated that the quality of the virtual reality graphics effected empathy, a lack of bystander intervention options reduced effectiveness, and customization cues had little effect on participants. Results from the laboratory experiment and follow-up focus group study suggest that in some cases, it may be difficult to use VR simulations to elicit empathy-related prosocial responses. While further study is needed to clarify what features of VR simulations might make them most effective in encouraging certain prosocial behaviors, findings here suggest that photorealistic graphics should be used in VR simulations to evoke empathy, additional intervention capabilities in VR simulations could make them more effective in producing bystander intervention behavior, and that customization cues should be prominent and possibly individually tailored. '