Three hundred and fifty-three Pittsburgh school children were randomly assigned to nine classes taught by the basal approach and nine classes taught by the integrated experience approach. Five readiness measures were administered at the beginning of the first grade, and the Stanford Achievement Test was given in May of first, second, and third grades. Analysis of the data yielded the following findings: (1) significant relationships existed between the predictor variables and the criterion variables of reading achievement; (2) the Murphy Durrell Readiness Test and the Metropolitan Readiness Tests were the strongest contributors to prediction; (3) when unique subtests of the two tests were combined, the combination of the Word Meaning subtest of the Metropolitan with the three subtests of the Murphy Durrell resulted in the highest prediction-achievement correlation; (4) correlation between prediction and achievement did not drop significantly from grade to grade; (5) sex, mental age, and instructional method did influence the prediction accuracy; (6) the Pintner Cunningham Primary Test made a moderate contribution to prediction; and (7) the Banham Checklist and Thurstone Jeffrey Tests were not adequate predictor instruments. References are included. (AW)