Over three decades of research have established a positive connection between fantasizing about aggression and enacting aggression. Such findings have provided strong evidence against the catharsis view of aggressive fantasy. However, little attention has been paid to the potentially nuanced nature of the link between fantasy aggression and actual aggression. In the present article, we examined the influence of four variables in the aggressive fantasy-aggressive behavior link: gender, exposure to violence, fantasy absorption, and level of fantasy about harm befalling loved ones and the self (dysphoric fantasy). Using data from a diverse, community-based sample of 7-14-year olds and their mothers, we replicated the general finding that aggressive fantasy is positively associated with real-world aggressive behavior. However, we also found that the interaction of aggressive fantasy and exposure to violence related significantly to aggression, as did the relation between aggressive fantasy and dysphoric fantasy. When exposure to violence was low, even high levels of aggressive fantasizing did not predict aggressive behavior, and, when aggressive fantasizing was low, even high levels of exposure to violence did not predict aggressive behavior. Similarly, when dysphoric fantasy was high, the connection between fantasy aggression and real aggression was markedly attenuated. The implications of these findings for intervention efforts and future research are considered.