STUDIES INVESTIGATING THE LEVEL OF ACHIEVEMENT OF INTELLECTUALLY GIFTED UNDERACHIEVING PUPILS IN THE ELEMENTARY SCHOOL WERE MADE BETWEEN TWO GROUPS OF UNDERACHIEVERS. ONE GROUP WAS PLACED IN A CLASS OF HIGH ACHIEVERS (HOMOGENEOUS) WHILE THE OTHER GROUP WAS PLACED IN A HETEROGENEOUS CLASS. IT WAS HYPOTHESIZED THAT UNDERACHIEVERS IN THE HOMOGENEOUS CLASS WOULD BE STIMULATED TO RAISE THEIR LEVEL OF ACHIEVEMENT, WOULD BECOME MORE CREATIVE, AND WOULD PERCEIVE THEIR PEERS AND PARENTS AS BEING MORE ACCEPTING OF THEM. INTELLECTUALLY GIFTED CHILDREN IN GRADES TWO THROUGH FIVE WERE SELECTED ON THE BASIS OF TEACHER REFERRALS, GROUP ACHIEVEMENT AND INTELLIGENCE TESTS, AND THE VOCABULARY SCALE OF THE 1937 STANFORD-BINET TEST. THESE CHILDREN WERE INVOLVED IN THE EXPERIMENT FOR 2 OR 3 YEARS. NOTHING WAS ADDED TO THE REGULAR CURRICULUM FOR EITHER GROUP. THE MEASURING SERVICES USED WERE THE CALIFORNIA ACHIEVEMENT TESTS. THE PERCEIVED PARENT ATTITUDE SCALE, THE CALIFORNIA TEST OF PERSONALITY, AND TWO CREATIVE ABILITY TESTS FROM GUILFORD. RESULTS SHOWED THAT THE UNDERACHIEVING GIFTED PUPILS PLACED IN HOMOGENEOUS CLASSES MADE SIGNIFICANT GAINS IN APPROACHING THEIR EXPECTED LEVEL OF ACADEMIC PERFORMANCE. HOWEVER, HIGHER GAINS IN CREATIVITY WERE ONLY PARTIALLY SUPPORTED. ALSO THE HYPOTHESIS THAT THESE PUPILS IN HOMOGENEOUS CLASSES WOULD BE BETTER ACCEPTED BY THEIR PEERS WAS NOT SUPPORTED.