How does armed conflict affect attitudes that tolerate violence against women? This article examines the effects of armed violence on preferences for punishing crimes against women using original quantitative data from 80 focus groups across 20 villages in Democratic Republic of Congo and a matched pair design. Challenging unidirectional logics within theories of violence against women, the data reveal that local exposure to armed violence increases how severely men prefer to punish rape while decreasing how severely men and women prefer to punish domestic violence. Building inductively, I develop a theory of protective masculine norms to account for armed conflict's gendered and crime-specific effects. When armed violence heightens demand for local male protection, crimes perceived to pose a community threat are affected differently from "private" crimes.Verification Materials: The data and materials required to verify the computational reproducibility of the results, procedures, and analyses in this article are available on the American Journal of Political Science Dataverse within the Harvard Dataverse Network, at: https://doi.org/10.7910/DVN/LBAV1Q.A large literature has demonstrated that harmful gender attitudes and practices emerge when armed group members organize, bond, and engage in war (Cohen 2013;Morris 1996;Wood 2018;Zurbriggen 2010). Studies also demonstrate that these harmful attitudes and practices spill over to the civilian population and unhinge social norms and sanctions that keep everyday violence perpetrated against women at bay (Haglund and Richards 2018; Østby, Leiby, and Nordås 2019; Zurbriggen 2010).Does armed conflict lead to more tolerant attitudes toward violence against women? Although the above-mentioned studies suggest increased attitudinal tolerance of violence against women, other studies have found that women can take on newly empowered roles during civil war (Berry 2018;Lazarev 2019;Wood 2008) with potentially opposite implications.This article examines the effects of community exposure to armed group violence on preferences for punishing violence against women in eastern Democratic Republic of Congo (DR Congo). I study preferences for punishing two forms of violence against women, rape (specifically, nonintimate partner rape) and domestic violence (specifically, wife beating), and focus on the mechanism of civilian self-protection efforts to explain attitudinal change.The findings show that men prefer to punish perpetrators of rape within the community more severely after exposure to armed violence. In contrast, both men and women prefer to punish perpetrators of domestic violence less severely after exposure to armed violence. These results are inconsistent with microlevel implications of established logics within theories of violence against women.Building inductively, I offer an alternative explanation: a theory of protective masculine norms. The theory begins with the need for community self-protection in the immediate aftermath of armed violence and then