A STUDY WAS CONDUCTED TO FIND OUT WHETHER OR NOT A SIGNIFICANT CHANGE IN PUPILS' ACADEMIC POTENTIAL (AS MEASURED BY A STANDARDIZED GROUP TEST OF INTELLIGENCE) COULD BE MADE AS A RESULT OF APPROPRIATE MODIFICATIONS OF THE TEACHING-LEARNING ENVIRONMENT TO MEET THE SPECIFIC NEEDS OF DISADVANTAGED PUPILS. DEPRIVED NEGRO PUPILS AND THEIR TEACHERS IN TWO NEW ORLEANS URBAN ELEMENTARY SCHOOLS TOOK PART IN AN EDUCATIONAL IMPROVEMENT PROJECT, INVOLVING A SPECIAL INSTRUCTIONAL PROGRAM. PRELIMINARY TESTING SUPPLIED DEMOGRAPHIC DATA ON BOTH TEACHERS AND PUPILS, NONCOGNITIVE DATA ON TEACHER VARIABLES, AND COGNITIVE DATA ON PUPILS. AT THE PROJECT'S INCEPTION, PUPILS HAD SUCH LEARNER BEHAVIOR CHARACTERISTICS AS DEPRESSED LEARNING POTENTIAL, LOW READINESS LEVELS, INADEQUATE READING SKILLS AND POOR IN-SCHOOL ACHIEVEMENT. THE LONG FORM OF THE CALIFORNIA TEST OF MENTAL MATURITY WAS GIVEN TO ALL (APPROXIMATELY 2,200) PUPILS IN THE FALL OF 1966, PRIOR TO THE INTERVENTION PROGRAM. INTERIM POSTTESTS WERE GIVEN IN THE SPRING OF 1967, AND POSTTESTS WILL BE GIVEN ANNUALLY IN GRADES ONE, THREE, AND FIVE. THERE WERE 305 GRADE ONE, 205 GRADE THREE, AND 250 GRADE FIVE PUPILS PARTICIPATING. ALL GAINS BETWEEN THE PRETESTS AND POSTTESTS IN GRADES ONE, THREE, AND FIVE WERE STATISTICALLY SIGNIFICANT AT OR BEYOND THE .001 LEVEL EXCEPT THE GRADE FIVE NONLANGUAGE GAIN, WHICH WAS SIGNIFICANT AT THE .05 LEVEL. LANGUAGE GAINS WERE HIGHER IN ALL CASES THAN NONLANGUAGE GAINS. THESE PRELIMINARY RESULTS APPEAR TO INDICATE THAT IQ SCORES, AS AN INDEX OF ACADEMIC POTENTIAL, CAN BE IMPROVED SIGNIFICANTLY WHEN GENERAL LEARNING ENVIRONMENT MODIFICATIONS ARE COUPLED WITH INSTRUCTIONAL INTERVENTION PROGRAMS. A DISTINCTION NEEDS TO BE MADE BETWEEN STATISTICAL AND EDUCATIONAL SIGNIFICANCE IN USING DATA TO PLAN OR EVALUATE PROGRAMS. (MS)