An alarming 30 percent of college freshmen in the nation are considered unprepared for college coursework and assigned to remedial education upon entry. Recent studies have explored the effects of remediation on college outcomes using rigorous quasi-experimental designs, but the results of these studies are inconclusive and the designs typically focus only on students on the margin of preparedness. Moreover, little is known about the extent to which students actually comply with remedial placement, and variation in compliance and timing of remediation may explain the mixed findings in prior impact studies. This dissertation uses administrative data from Florida to thoroughly examine students' coursework in remedial mathematics, reading, and writing, and the effects of such coursework on performance at the first college-level course. Specifically, I answer the following questions: (1) Does passing remedial courses improve performance at the first college-level course? (2) Does the effect of passing remedial courses vary by the student's level of preparedness? (3) Does starting remedial work in the first term of college affect the impact of passing remedial courses? Findings: At the state level, passing math and reading remediation does not seem to affect the probability of success in the first college-level math and reading courses, respectively. Passing writing remediation appears to help performance at college-level writing for students who started writing remediation after the first term of college. Positive cross-subject effects of reading and writing remediation were found, but these effects were not robust across specifications. For all three subjects, there is suggestive evidence of effect heterogeneity across institutions. Throughout the analyses, positive effects, when detected, were generally modest. Remedial courses seem to associate with little impact on performance at college-level courses.