We conducted a meta-analysis of 36 studies of Reading Recovery (RR), an intensive tutorial intervention designed to develop the literacy skills of low-performing first-grade students. Few individual studies of the program have yielded conclusive evidence regarding the program's effectiveness due to various methodological limitations. We relied on specific meta-analytic strategies to combine as much available evidence as possible to study overall program effects. We also analyzed the results from the few more rigorously designed studies separately. In general, we found positive program effects for both discontinued and not discontinued students on outcomes tailored to the program and standardized achievement measures. RR effects were most pronounced, however, for discontinued students on measures designed for the program. Contrary to conventional belief, we found no evidence suggesting that prior observed effects could be explained completely by factors resulting from methodological flaws (e.g., regression artifacts). (Contains 4 tables.)