The existing research on the appeal of media violence has led to an apparent incongruity: violent content tends to increase selective exposure to media, but violence often decreases enjoyment. In this experiment, we used two independent manipulations to assess the role of violence in both selective exposure and enjoyment in order to examine the relationship between the two. Program descriptions for four prime-time television dramas were altered to create violent and nonviolent descriptions for each episode. Then the episodes themselves were edited to create violent and nonviolent versions of each. Participants (N = 191) were more likely to choose violent descriptions to watch, but enjoyed the nonviolent episodes more than the violent episodes. Moreover, the nonviolent episodes were rated as more enjoyable even when the participants had chosen to watch a violent program description. From a theoretical perspective, these results suggest the need to move beyond explaining the appeal of violence in terms of increased enjoyment and instead further explore other motivations that could be driving selective exposure to violent content.