This review examined the literature on peer tutoring in reading with students with disabilities, using the methodology of best evidence synthesis. This approach combines positive aspects of meta analysis with traditional integrative review procedures. It requires all studies reviewed to meet stringent a priori standards. Analysis of the 11 studies in this review indicates that peer tutoring in reading with students with disabilities can be effective. Peer tutoring was found to have an overall effect size of instruction students with disabilities typically receive. However, it was not found to be more effective than teacher-led instruction when the teacher implemented another research intervention. Average effect sizes across dimensions were consistent; however, individual treatments evidenced variable effect sizes. Treatments in which students with disabilities were paired with normally achieving peers and allowed to serve in the role of tutor some of the time consistently produced strong effect sizes and significant findings. An appendix lists the methodological reasons that 19 additional studies were excluded from analysis. Another appendix details the coding process and compares the 11 selected studies in detail. (94 references) (DB)