This article describes a randomized evaluation of The Reading Edge, a reading program for middle school students. The Reading Edge was designed to integrate findings of research on cooperative learning and metacognitive reading strategies into a replicable reading instructional package that could be implemented effectively in Title I middle schools. In this study, 405 sixth graders in two high-poverty, rural middle schools previously unfamiliar with the program were randomly assigned to participate in The Reading Edge or to continue with their existing reading programs. After one year of instruction, observations of classroom use of metacognitive strategies, cooperative learning, goal setting/feedback, and classroom management, showed moderate levels of implementation in Reading Edge classes but little use of metacognitive strategies, cooperative learning, or goal setting/feedback in control classes. Statistically significant differences in student scores on the Vocabulary subscale of the Gates-MacGinitie Reading Test, and marginally significant scores on the Total Achievement score, provide support for the basic reading model, but larger and longer studies are needed to establish the full effects of this approach. (Contains 1 table.)