There is increasing evidence that the first few years after birth are particularly important in child development and present opportunities for enrichment but also vulnerabilities do to poverty and other social stressors. Elected officials have begun proposing potentially costly programs to intervene early in the lives of disadvantaged children. Have such interventions been demonstrated to yield substantial benefits? To what extent might they pay for themselves through lower welfare and criminal justice costs incurred by participating children as they grow into adults? This study synthesizes the results of a number of previous evaluations in an effort to answer those questions. Conclusions are that under carefully controlled conditions, early childhood interventions can yield substantial advantages to recipients in terms of emotional and cognitive development, education, economic well-being, and health. (The latter two benefits apply to the children's families as well.) If these interventions can be duplicated on a large scale, the costs of the programs could be exceeded by subsequent savings to the government. However, the more carefully the interventions are targeted to children most likely to benefit, the more likely it is that savings will exceed costs. Unfortunately, these conclusions rest on only a few methodologically sound studies. The authors argue for broader demonstrations accompanied by rigorous evaluations to resolve several important unknowns. These include the most efficient ways to design and target programs, the extent to which effectiveness is lost on scale-up, and the implications of welfare reform and other "safety net" changes.