This series of notes summarizes findings of a project entitled "What development interventions work?" undertaken by researchers of the Chronic Poverty Research Centre, the International Food Policy Research Institute, and Data Analysis and Technical Assistance Ltd. As part of a larger longitudinal study that resurveyed 1,907 households and 102 villages in 14 of Bangladesh's 64 districts, the project focused on assessing the long-term impacts of a number of anti-poverty interventions-specifically, microfinance, agricultural technology, and educational transfers- on a range of monetary and nonmonetary measures of well-being. This note focuses on the long-term impacts on men's and women's assets of disseminating agricultural technologies to individuals compared with groups. It is hoped that these results will help policymakers, donors, and other stakeholders to effectively evaluate different interventions thereby contributing to the design of future anti-poverty programs in South Asia. Many of the best-studied programs targeting women in Bangladesh- particularly microfinance programs directed toward poor women-have operated through women's groups. In these programs, group liability acts as a substitute for personally owned assets as a form of collateral. Whereas some evidence does suggest that collective action has a positive impact on gender relations and broader development objectives like reducing poverty, when evaluating impact, many studies do not satisfactorily account for other factors associated with participation in collective action. For example, it is possible that women who are more "empowered" to begin with are more likely both to participate in and benefit from collective-action programs, perhaps because of greater wealth, higher levels of schooling, or better social connectedness. The panel data set employed in this study addresses this issue by providing the necessary conditions for more rigorous, long-term impact evaluation.