The purpose of the 1990-91 Reading Recovery Program in the Columbus, Ohio public schools was to provide early intervention to underachieving first-grade pupils who appeared unlikely to read successfully without intensive instruction. The program featured individualized one-on-one lessons provided by specially trained teachers. The lessons were based upon diagnostic instruments designed to provide a comprehensive assessment of the pupil's development of reading and writing strategies. Daily 30-minute individualized lessons included a variety of instructional activities designed to help pupils develop reading strategies. The major evaluation effort was accomplished through administration of Metropolitan Achievement Tests and locally constructed instruments. The program served 297 pupils in 38 elementary schools, with an average pupil enrollment of 64.8 days. Of this group, 139 students were included in the evaluation sample. Results indicated that the three desired outcomes were met: (1) 69.8% of the pupils had a normal curve equivalent (NCE) gain of 3.0 or better; (2) 96.1% of the pupils were not retained in grade 1; and (3) 98.1% of the pupils had parental involvement. Results also indicated that 80.8% of the teachers believed the program was of value to their pupils, and that the program was expensive in comparison to other compensatory programs. Findings support the continuation of the program. (Eight specific recommendations and 23 tables of data are included; survey instruments, record-keeping sheets, and comments from the classroom teacher survey are attached.) (RS)