Each year , thousands graduate high school academically underprepared for college. Many must take remedial or developmental postsecondary coursework, and there is a growing debate about the effectiveness of such programs. This paper examines the effects of remediation using a unique data set of over 28,000 students. To account for selection biases, the paper implements an instrumental variables strategy based on variation in placement policies and the importance of proximity in college choice. The results suggest that students in remediation are more likely to persist in college in comparison to students with similar backgrounds who were not required to take the courses. (Contains 7 tables and 32 footnotes.)