This article uses data from the Project on Human Development in Chicago Neighborhoods (PHDCN) to investigate the sex gap in violent crime. We posit that the sex gap decreases at higher levels of exposure to violent peers, and we test two competing explanations for this possibility. The first hypothesis, based on previous research using the PHDCN, suggests that there is a stronger nonlinear relationship between exposure to violent peers and respondent violence for females than for males. The second hypothesis points to sex differences in the shape of the nonlinear relationship between peer violence and respondent violence. We find that a decrease in the sex gap at higher levels of peer violence is due to (a) a stronger effect of peer violence exposure on individual violence for females than for males, and (b) a nonlinear peer violence/respondent violence relationship in the form of a decelerating curve that is consistent across the sexes.