There is growing evidence that the current method of identifying students with a learning disability (LD) is ineffective. The wait-to-fail model of assessing students after second/third grade as well as conceptual problems in using intelligence tests for identification result in students not receiving the assistance they need during the early-elementary school years. The educational community is discussing response to intervention (RTI) as an alternative assessment method. This study explored the assessment components of the Reading Recovery (RR) program as part of a future RTI model. By means of a discriminant-function analysis, a retrospective study of third- through fifth-grade students who participated in RR during first grade investigated assessment elements (beginning text level, ending text level, and number of weeks' participation in RR) of this program. The results showed that RR assessment elements were significant predictors of first-grade students who were later identified as having a reading disability. Ending text level was consistently the largest predictor of students later classified as having a reading disability or not. (Contains 4 tables.)