This article evaluates the impact and cost-effectiveness of offering an innovative middle school model--the Sistema de Aprendizaje Tutorial (SAT)--to Honduran villages instead of traditional middle schools. We identified a matched sample of villages with either type of school and collected baseline data among primary school graduates eligible to enroll in middle schools. After 2 years, the test scores of children residing in SAT villages were 0.2 standard deviations higher than children in other villages, though the per-student cost in SATs was at least 10% lower than traditional schools. The article is one of the few studies to rigorously evaluate a scaled-up instructional reform in a poor country, implemented with an alternative model of teacher recruitment and contracting.