Existing evidence on the effectiveness of conditional transfers leaves a number of open questions. In this study, we utilize the changes in the coverage of Bangladesh’s Primary Education Stipend Program (PESP), a conditional small transfer to the mothers of eligible children, to investigate three of these open questions using longitudinal data from the Bangladesh Integrated Household Survey. First, we examine if a transfer as small as PESP, which is unlikely to make large distortions in decisions, can have a positive impact on household welfare. We find that the transfers improved education and health outcomes for the children and increased households’ protein consumption. Second, motivated by the findings around the intra-household disparity in the effect of transfers, we focus on welfare outcomes for women alone. We find that the transfers improve female minor asset holdings, employment, and decision-making power, and reduce verbal abuse from intimate partners, countering narratives that express concern over increases in intimate partner violence. At the same time, we find an increase in domestic workload for women, reducing their measures of life satisfaction. Third, we compare the effect of receiving a transfer to that of losing the stipend. While largely symmetric, some effects of receiving the transfers on female empowerment remain when they no longer receive the stipend. Our findings affirm the potential for positive impacts even with small cash transfers but highlight the increased burden of targeting women as transfer recipients.