This study investigated the relationship between measured intelligence and discrepancies existing between actual achievement levels and projected achievement levels in mildly handicapped resource room students in Arkansas. Twenty-seven students had been identified as learning disabled and 19 as mildly mentally retarded. Subjects each received approximately 1 hour per day of remedial instruction in a resource room setting. Subjects were individually administered the Wechsler Intelligence Scale for Children--Revised and the Wide Range Achievement Test--Revised prior to placement in the resource room program and at the completion of approximately 3 years in the program. Major findings included: the mean intelligence quotient did not change significantly in the learning-disabled group though it changed in a positive direction for the mentally retarded group; discrepancies decreased between groups with remediation in reading and spelling; discrepancy in arithmetic between the two groups increased, with the mentally retarded group showing less discrepancy in arithmetic, the only area of significant improvement for either group; for the total group, reading achievement was the area most affected by remediation, but in a negative manner; and for the total group, variability in scores was greatest in arithmetic achievement. Appended are the formulae used in the regression model computer program. Includes five references. (DB)