Our paper utilizes variation across the 50 U.S. states to examine the relationship between public expenditures on children and child outcomes. We find that public expenditures on children are related to better child outcomes across a wide range of indicators, including measures of child mortality, elementary-school test scores, and adolescent behavioral outcomes. States that spend more on children have better child outcomes even after taking into account potential confounding influences. Our results are robust to numerous variations in model specifications and to the inclusion of proxies for unobserved characteristics of states. Our sensitivity analyses suggest that the results we present may be conservative, yet our findings show that public investments in children yield broad short-term returns in the form of improved child outcomes.