Program providers work extremely hard to offer services that they believe will enhance the lives and future prospects of children and youth. But how can they be sure that their program is having the intended impacts? How can funders and policy makers be sure that their investments in programs are paying off in substantially better outcomes for children and youth? How can parents and children know that they are choosing to be involved in a worthwhile activity? Random assignment studies are only one kind of research, and they are not appropriate for every situation. But when the question is "What works?" and rigorous evidence about cause and effect is needed, a random assignment experimental evaluation can be very useful. In addition, a random assignment evaluation can provide credible evidence about the size of impacts found for the program participants. This brief describes the importance of conducting random assignment evaluation studies, examines how they are done, explains the differences between biomedical and social experiments, and addresses reservations that program providers may have. (Contains 10 notes.)