Researchers from IPA, along with CARE staff and their implementing partners, conducted a randomized evaluation of Village Savings and Loans Association (VSLA) programs in Ghana, Malawi, and Uganda to examine two questions: Who joins savings groups? And, what is the impact on households from programs that promote savings groups? The evaluation used a randomized control trial (RCT) design, in which eligible communities were randomly divided into two sets: a set of villages with access to a VSLA program (the treatment group) and a set of villages where the program was not implemented during the study (the control group). The study started in Ghana in 2008 and in Malawi and Uganda in 2009, and the final data collection took place in 2011 in the three countries. Each site included a panel survey in which households were surveyed before the start of the program implementation and again two or three years later. Over 15,000 households in almost 950 communities were surveyed. The surveys covered a large variety of topics, including health, education, income-generating activities, asset holdings, food consumption, non-food expenditure, intra-household decision making and community involvement. At the time of the endline survey, after an average of two years of program implementation in the three sites, one third of respondents had joined a VSLA group. On average, members had been part of a group for 15 months and 61% of members had gone through a full savings cycle, normally lasting between 8 and 12 months. The evaluation should thus be thought of as assessing the relatively short-term impacts of the intervention. [62 pages]