In this article, I examine the current policy of full inclusion of children receiving special education services in regular education classrooms. Specifically, I focus on the policy's effects on the "classmates" of children with special needs, with a particular focus on classmates of students with serious emotional problems. Results suggest that students with a classmate with a serious emotional problem experience reductions in first-grade test scores, especially students in low-income schools. Results that attempt to capture sorting across and within schools using school-level fixed effects specifications are qualitatively similar. The magnitude of the reduction in mathematics achievement is approximately 30-60 percent of the size of the adjusted black-white achievement gap. Since nearly 10 percent of the student population has a classmate with a serious emotional problem, the aggregate effect on test scores of the policy of including these students is potentially quite large.