Evaluated the Wechsler Intelligence Scale for Children-Revised (WISC-R) as a tool for understanding children's learning problems. Analyzed the predictive utility of several WISC-R subtest recategorizations with regard to academic achievement for 105 children referred for psychoeducational evaluation because of classroom learning problems. Stepwise regression analysis indicated that many of the regroupings were significant predictors of academic achievement. (Author/KS)