The relative effectiveness of two computer-assisted instructional programs designed to provide instruction and practice in foundational reading skills was examined. First-grade students at risk for reading disabilities received approximately 80 h of small-group instruction in four 50-min sessions per week from October through May. Approximately half of the instruction was delivered by specially trained teachers to prepare students for their work on the computer, and half was delivered by the computer programs. At the end of first grade, there were no differences in student reading performance between students assigned to the different intervention conditions, but the combined-intervention students performed significantly better than control students who had been exposed to their school's normal reading program. Significant differences were obtained for phonemic awareness, phonemic decoding, reading accuracy, rapid automatic naming, and reading comprehension. A follow-up test at the end of second grade showed a similar pattern of differences, although only differences in phonemic awareness, phonemic decoding, and rapid naming remained statistically reliable.