Eighteen first grade children (10 boys and eight girls, IQ's 91 to 128) who evidenced a developmental deficit on the Sapir Developmental Scale were placed in one of two groups. Twelve children in an experimental group compromised a self contained class and were given deficit centered training. The six in the control group were placed with 12 children without problems and given the traditional curriculum without deficit centered training. In the deficit centered program emphasis was placed on sensory stimulation in a carefully planned environment. The experimental group did significantly better on many, but not all of the intellectual, perceptual, and language tests; but data on academic achievement failed to show significant differences. Two factors are noted which might have contributed to the results: children who develop unevenly may have a distinctive learning pattern and process information differently, and neurological impairment could have a negative effect on academic performance regardless of WISC IQ. These factors and questions raised by the study are discussed. (RJ)