We study the peer effects of grade retention in the context of Indiana's statewide third-grade retention policy. When a retention occurs, it changes the peer group for two cohorts: rising fourth graders who lose a peer and rising third graders who gain a peer. We identify peer effects in both cohorts by leveraging plausibly exogenous variation in cohort-level retention rates caused by a discontinuity in the rule that determines which students are retained. We find that when a student is retained, rising fourth graders who lose the student as a peer are less likely to have a disciplinary incident. However, the effect fades out after one year, and there is not consistent evidence of peer effects on other outcomes for promoted students. For rising third graders who gain the retained student as a peer, we find no evidence of peer effects. We conclude that the total peer effects of grade retention are limited. [This paper was supported by Notre Dame's Center for Research on Educational Opportunity (CREO), the Institute for Educational Initiatives, and their partnership with the Indiana Department of Education.]