A variety of theories of skill formation suggest that investments in schooling and other dimensions of human capital will have lower returns if children do not have adequate levels of cognitive and social skills at an early age. This paper analyzes the impact of a randomized cash transfer program on cognitive development in early childhood in rural Nicaragua. It shows that the program had significant effects on cognitive outcomes, especially language. Impacts are larger for older pre-school age children, who are also more likely to be delayed. The program increased intake of nutrient-rich foods, early stimulation, and use of preventive health care-all of which have been identified as risk factors for development in early childhood. Households increased expenditures on these inputs more than can be accounted for by the increases in cash income only, suggesting that the program changed parents' behavior. The findings suggest that gains in early childhood development outcomes should be taken into account when assessing the benefits of cash transfer programs in developing countries. More broadly, the paper illustrates that gains in early childhood development can result from interventions that facilitate investments made by parents to reduce risk factors for cognitive development.