This paper measures the effects of collegiate class size on college retention and graduation. Class size is a perennial issue in research on primary and secondary schooling. Few researchers have focused on the causal impacts of collegiate class size, however. Whereas college students have greater choice of classes, selection problems and nonrandom sorting make it difficult to estimate causal impacts. Using unique data and exogenous variation in class size, we estimate the impacts of class size using a sample of nearly 60,000 four-year college students. Using an instrumental variables approach to control for selection bias, the results suggest an increase in collegiate class size leads to an increase in dropout rates and a reduction in on-time degree completion, but no change in long-run degree completion.