States and districts are revising discipline policies to reduce out-of-school suspensions (OSSs), but the consequences of these reforms are largely unknown. We examine a reform in Philadelphia that prohibited OSS for classroom disorder infractions. Employing a difference-in-differences approach, we examine the relationship between the reform and student suspensions, achievement, and attendance. For students suspended before the reform, classroom disorder OSS decreased and attendance (but not academic achievement) improved following the reform. Postreform changes in peer outcomes varied with school-level implementation: in schools that eliminated classroom disorder OSS, peer math achievement and attendance were unaffected, whereas peer math achievement declined and attendance decreased in schools that did not fully implement the district-level reform.