This paper evaluates a reading tutoring program—funded by the Chilean Education Ministry—in which fourth-grade students in forty-five randomly selected low-performing schools receive, over the span of three months, fifteen tutoring sessions with volunteer college students. Program effects on reading test-scores and attitudes are strongest—between 0.2 and 0.3 standard deviations—in high-poverty schools and in schools that complied with the fifteen-session cycle and managed to maintain small tutoring groups.