Some researchers argue that growing women's autonomy has worsened homicide victimization rates, attributed to a backlash effect. Others argue that increased women's autonomy has an ameliorative effect, lessening homicide rates. To address this ambiguity, this article puts forward both methodological and theoretical modifications. First, data (from the UNODC Global Study on Homicide) is analyzed longitudinally rather than cross-sectionally. This analytical focus on changing trends is more appropriate to test existing theories. Second, we use measures of women's autonomy across a number of social institutions, estimating and comparing effects in terms of women's gains in political office, the labor market, and for adolescent fertility. Women's gains in political office have no relationship with homicide rates. Integration of women in the labor market shows a curvilinear relationship, demonstrating a backlash affect until women are about 40% of the labor market, when homicide rates start declining. However, for low-income countries the labor market integration shows a linear, positive effect on homicide rates. Adolescent fertility has a strong, positive association with homicide trends-in fact, among the strongest of all predictors. Finally, gains in women's autonomy are similarly consequential for the homicide against males and females. Overall, findings show that different measures of autonomy have distinct effects, and that gains amongst a restricted political elite do not necessarily spill over to other social institutions, such as the private sphere of families, which can be more impactful for changes in homicide rates.