Prior to initial registration at most community colleges, students typically take standardized placement exams in college-level reading, writing, and math. More than half of community college students nationally are assigned to remedial education as a result. Prior research has found mixed evidence, at best, of the impact of remediation on future outcomes. This paper builds on the prior literature in three ways. First, we specify three alternative models of the impact of remediation: remediation as development, diversion, or discouragement.  Second, we are able to look at two important outcomes that have not been examined in the previous literature: 1)
impacts of remedial assignment on the initial decision to matriculate, and 2) impacts of remedial assignment on student learning, as measured by a subsequent (and entirely separate) standardized exam. Finally, we exploit the richness of our dataset to examine heterogeneity in the
impact of remedial assignment by prior predicted dropout risk as well as for a sample of students we identify as misplaced into remediation. Our findings suggest that assignment to remediation does not effectively develop skills needed for college-level success, but neither does it appear to be a significant discouragement to student progress, at least for those near the remedial cutoffs. Instead, the primary effect of remediation appears to be diversionary: students enroll and persist at the same rates but simply take remedial courses instead of college-level courses. These diversionary effects are larger for the lowest-risk students. We do, however, find strong
evidence of both diversionary and discouragement effects for a subsample we identify as wrongly assigned into remediation.