A growing literature establishes that high-quality early childhood interventions that enrich the environments of disadvantaged children have substantial long-run impacts on a variety of social and economic outcomes. Much less is known about their effects on health. This paper examines the long-term health impacts of two of the oldest and most widely cited U.S. early childhood interventions evaluated by the method of randomization with long term follow-up: the Perry Preschool Project (PPP) and the Abecedarian Project (ABC). We document that the boys randomly assigned to the treatment group of the PPP have significantly lower prevalence of behavioral risk factors in adulthood compared to those randomized to the control condition, while those who received the ABC intervention enjoy better physical health. Estimated effects are much weaker for girls. Our permutation-based inference procedure accounts for the small sample sizes of the ABC and PPP interventions, for the multiplicity of the hypotheses tested, and for nonrandom attrition from the panel follow-ups. We conduct a dynamic mediation analysis to shed light on the mechanisms producing the estimated treatment effects. We document a significant role played by enhanced childhood traits, above and beyond experimentally enhanced adult socioeconomic status. Overall, our results show the potential of early life interventions for preventing disease and promoting health.