How does women's engagement in civil society organizations (CSOs) differ from that of men, and what factors predict women's willingness to hold the state accountable? We analyze these questions in the context of rural and urban Mali, leveraging face-to-face data collected as part of a civil society mapping project during February -- March 2020 and December 2020, and an in-depth survey conducted with leaders from a randomly-selected subset of these CSOs during January -- March 2021. First, we explore the characteristics of women's groups compared to other CSOs. Second, we explore their likelihood of sanctioning a hypothetical corrupt mayor. We use an embedded survey experiment to try to understand these groups' willingness to report on the mayor as well as the payments they expect to receive from the mayor in order to not share information about the mayor's corruption. We find that women in Mali are often highly organized at the local level---frequently in self-help groups or organizations related to gendered economic activities. However, these more economically-focused groupings of Malian women frequently do not translate into civic activity. They are not typically recognized by outside actors as viable CSOs that could incentivize better governance; their strong networks and group infrastructure represent untapped social capital. We also find that CSOs comprised of women have lower informational and technical capacity, especially due to lower levels of political knowledge, and incur a higher cost of sanctioning public officials---though they have greater mobilization capacity. Women's engagement in decision-making in the home also predicts one's CSO being more hierarchical and having greater technical capacity, as well as having higher expected transfers from the mayor---findings that do not vary with respondent or CSO gender. Turning to the extent to which women involved in CSOs are willing to sanction the state, we find that they are generally less willing than men to sanction, but become more likely when their CSO is less hierarchical, when their technical capacity is higher, and when their political knowledge is greater. However, priming their importance as a CSO (by telling them they were identified by well-connected citizens as being influential) actually reduces sanctioning---perhaps by making them fear that those recommending them will engage in reprisals. Overall, our findings provide useful evidence on the organizational capacity of women and how it is likely to affect the state.