Districts are an important unit for administrative purposes, but they vary little in their impact on students' attainment, at least in the UK. Further, government attempts to raise attainment are often disappointing. The project described in this article aimed to engage schools in reform to change students' attainment and attitudes in schools across a whole district. The intervention, peer tutoring, has a good research pedigree in small-scale studies, but scaling it up to district-level implementation has not been rigorously evaluated. Over 2 years, 129 elementary schools in 1 Scottish district were randomly assigned to different interventions. The implementation was not perfect, but the results were positive with respect to cross-age tutoring, which had effect sizes of about 0.2. Despite limitations, the study demonstrates that it is possible to carry out a clustered randomized controlled trial (RCT) on a large scale working with districts and suggests that peer tutoring has promise when scaled up. (Contains 10 tables and 1 figure.)