A remedial reading program designed for intermediate-grade students who read from 1 to 7 years below grade level was studied. The program provided individualized instruction within classes homogeneously grouped on the basis of reading level only. Six seventh-grade classes were studied, with three acting as homogeneously grouped experimental classes and three acting as controls. All groups contained white, black, and Puerto Rican boys and girls, but sex and race were not factors in group establishment. Class sizes ranged from 22 to 28 students, and experimental students were transferred whenever reading levels changed. Materials were specially chosen to meet experimental class needs. Oral and silent reading, charts of progress, programmed materials, and skill developmental exercises were used in experimental classes. Pretests and posttests administered in September and June were analyzed using t-tests. All comparisons demonstrated significantly better gains in reading (p .01) for the experimental groups. This was true for all races and both sexes. It was concluded that homogeneous grouping was of considerable value in this remedial reading program for intermediate-grade students. Tables and references are included. (MS)