Pré-publication, Document de travail: Interventions targeting early childhood development hold promise for increasing human capital and reducing the intergenerational transmission of poverty. This paper presents results from a randomized evaluation of a preschool construction program in Cambodia, and suggests caution. The overall impact of the program on early childhood outcomes was small and statistically insignificant. For the cohort with highest program exposure, the impact on cognitive indicators was negative; with the largest negative effects among children of poorer and less educated parents. The results are consistent with frequent underage enrollment in primary school in the absence of preschools, stricter enforcement of the minimum age for primary school entry after the intervention, substitution between primary and preschool following intervention, and difference in demand responses to the new preschools between more and less educated parents. The results show that contextual and program specifics, and behavioral responses, can potentially lead to perverse effects of programs.