This paper experimentally examines the impacts of a largescale government program that increased the supply and quality of community preschools in rural Cambodia. The construction of new preschool facilities was paired with two demand-side interventions designed to stimulate additional enrollment into preschools. The newly constructed preschools caused an increase in enrollment rates but the demand-side interventions did not. One year after the program started, the paper finds small and significant impacts on cognitive (0.04 standard deviations) and socio-emotional development (0.09 standard deviations). The analysis shows that the cognitive impacts are driven by children from the wealthiest quartile, while the program had limited impacts on children from the poorest families. The effects on cognitive development increased after two years for the wealthiest (the cognitive gap widened) while the effects on socio-emotional development faded out across the board. Using detailed classroom surveys and in-class observations, the paper shows that the program had large impacts on the quality of preschool infrastructure and materials but only limited impacts the quality of educational processes—the results therefore suggest that further improvement of those processes might be needed to foster the development of disadvantaged children.