This analysis estimated the net impact of Reading Recovery® as it was conducted in Kalamazoo Public Schools for first graders in the 2003/04 cohort. Results suggest that Reading Recovery® has helped students who demonstrated difficulties with literacy skills at the beginning of the year to perform at levels much closer to their peers by the end of the year. Three comparison groups were used in the analysis: RR students only, all other first graders, and matched first graders. Analysis using all three comparison groups found evidence that students in Reading Recovery® raised their test scores more than students that were not in the program during the same period of time. These positive results were not found for all tests; for some of the tests the differences were statistically insignificant. In no case did researchers find a difference in favor of the comparison group that was statistically significant. Controlling for other characteristics available in the dataset, such as free and reduced price lunch status, gender and ethnicity does not alter the results. The analysis looked only at the net impact of the program and did not consider the cost effectiveness of Reading Recovery® nor did it compare the outcomes of Reading Recovery® to other reading programs.