Sixteen million children in developing and middle-income countries have been orphaned by HIV/AIDS, and at least another million children per year are rendered vulnerable by parental HIV/AIDS-related illness. Since 2003 the US government has provided approximately $1.6 billion to give four million of these children care and support through the President's Emergency Plan for AIDS Relief (PEPFAR). We conducted five studies to evaluate the effectiveness of PEPFAR's interventions for such children in East Africa and southern Africa. We found evidence of beneficial changes in school enrollment rates and on the psychosocial well-being of children. However, we could not demonstrate empirically the impact of most of the PEPFAR initiatives that we examined, primarily because of a lack of baseline data and clear outcome and impact indicators. We also found that many programs were spread so thin across a vulnerable population that little in the way of services actually reached beneficiaries, which raises questions about whether PEPFAR funds are sufficient, or if the program is attempting to do much with too few resources. We offer several recommendations, including better measuring the effect of programs for orphans and vulnerable children by collecting baseline data and conducting well-designed, rigorous outcome and impact evaluations.