TWO DIFFERENT USES OF VISUAL REPRESENTATIONS FOR SCIENCE INSTRUCTION WERE INVESTIGATED. THE FIRST STUDY EXAMINED THE HYPOTHESIS THAT USE OF PICTORIAL INSTRUCTION WOULD PRODUCE HIGHER CORRELATION BETWEEN RESULTS OF VISUAL APTITUDE TESTS AND LEARNING TESTS, AND THAT VERBAL INSTRUCTION WOULD PRODUCE HIGHER CORRELATION BETWEEN RESULTS OF VERBAL APTITUDE TESTS AND RESULTS OF LEARNING TESTS. TEST RESULTS SHOWED NO DIFFERENCE IN CORRELATION OF VISUAL APTITUDE OR VERBAL APTITUDE TEST RESULTS TO LEARNING ACHIEVED. APTITUDE MEASURES USED WERE (1) SPATIAL RELATIONS, (2) VERBAL REASONING, (3) ABSTRACT REASONING, AND (4) INTELLIGENCE. THE SECOND STUDY USED PICTORIAL REPRESENTATIONS IN REVIEW SESSIONS, COVERING STUDY OF MECHANICAL ADVANTAGE TO INVESTIGATE (1) WHETHER RETENTION WAS IMPROVED BY USE OF PICTURES IN REVIEW AND (2) WHETHER INCLUSION OF ADDITIONAL DIFFERENT EXAMPLES WOULD INCREASE RETENTION AND TRANSFER. RESULTS SHOWED THAT STUDENTS WHO REVIEWED BY PICTORIALLY PRESENTED MATERIALS HAD SIGNIFICANTLY BETTER RETENTION AND TRANSFER OF TRAINING THAN STUDENTS WHO DID NOT REVIEW AT ALL. ADDING NEW PICTORIAL EXAMPLES PRODUCED NO SIGNIFICANT RESULTS IN RETENTION BUT SHOWED BETTER TRANSFER WHEN ONLY THE ORIGINAL MATERIALS WERE REVIEWED. RETENTION AND TRANSFER WERE MEASURED 4 WEEKS FOLLOWING THE REVIEW SESSION. (AL)