Numerous initiatives by private philanthropies and the US government have supported school size reduction policies as an educational reform intended to improve student outcomes. Empirical evidence to support these claims, however, is underdeveloped. In this article, we draw on information from a longitudinal dataset provided by the Northwest Evaluation Association covering more than 1 million students in 4 US states. Employing a student fixed effects strategy, we estimate how a student's achievement changes as (s)he moves between schools of different sizes. We find evidence that students' academic achievement in math and reading declines as school size increases. The negative effects of large schools appear to matter most in higher grades, which is also when schools tend to be the largest.