A programmed-workbook cassette-delivered approach to letter and number recognition, "From Letters to Words," was compared with a traditional method of teaching letter recognition. The subjects were 153 disadvantaged first-grade students. The 75 experimental subjects were individually instructed by the investigators and teachers on the use of the program. Each experimental subject was assigned to work with the materials for 20 minutes each day. The 78 control-group subjects utilized the basal series approach supplemented by other teacher-supported materials. The analysis of the data revealed that the experimental group means were significantly higher on three subtests of the Gates Reading Readiness Test. The experimental group learned to distinguish between words, recognize lower and uppercase letters, and recognize numbers at a rate exceeding that of the control group. The differences were significant at the 1 percent level of confidence. No statistically significant differences were found between experimental boys' and girls' means. (Author/MKM)