Intended to improve language and reading skills of selected underachieving students, the 1987-88 Compensatory Language Experiences and Reading Program (CLEAR) of the Columbus, Ohio, public schools served 4109 pupils in grades 1-8 in 94 schools. The program featured individual and small group instruction arranged according to pupil needs, as well as projects utilizing computer-assisted instruction (CAI). The program was evaluated through administration of the Comprehensive Tests of Basic Skills as pretest and posttest. Test results indicated that of 2,349 pupils with good attendance records who were native speakers of English, average test score gain was 6.3 Normal Curve Equivalent (NCE) points for the 6.2 months of instruction, which met the program objective of 1.0 NCE per month of instruction. Analysis by grade revealed that the objective was met in grades 3, 4, 6, and 7. Average NCE gains for students using CAI were higher than those for the regular group in grades 1-5, and lower than those for the regular group at the middle school level. Recommendations include a review of selection procedures, instructional methods, class size, and test content to determine why pupils at some grade levels did not show desired growth. (Fourteen tables of data are included; appendixes include tables of data, and survey and evaluation instruments.) (SR)