THE OBJECTIVES OF THIS STUDY WERE TO DETERMINE--(1) IF THE NEGRO CHILD'S SELF-CONCEPT WOULD IMPROVE SIGNIFICANTLY WHEN HE LEARNS ABOUT HIS HERITAGE, (2) IF THE CAUCASIAN CHILD'S SELF-CONCEPT WOULD IMPROVE SIGNIFICANTLY BY A STUDY OF THE NEGRO'S HERITAGE AND CONTRIBUTION TO AMERICA, AND (3) IF COMMUNITY CONTACTS INFLUENCE CHILDREN'S SELF-CONCEPTS. ALL THE FOURTH-GRADE CHILDREN ATTENDING 26 INTEGRATED CLASSES IN GARY, INDIANA, WERE DIVIDED INTO--10 EXPERIMENTAL CLASSES, NINE EXPERIMENTAL CLASSES WITH TRANSPORTED STUDENTS, AND SEVEN CONTROL CLASSES. THE EXPERIMENTAL GROUPS STUDIED A UNIT ON AMERICAN NEGRO HISTORY. PRE- AND POST-TEST INSTRUMENTS USED WERE THE PIERS-HARRIS SELF-CONCEPT SCALE AND A TEST OF FACTUAL KNOWLEDGE ON NEGRO HISTORY AND CULTURE. DATA ON THE PRE- AND POST-TEST RESULTS WERE ANALYZED FOR SIGNIFICANT DIFFERENCES BY MEANS OF COVARIANCES. THE SELF-CONCEPT SCALE OF BOTH NEGRO AND WHITE CHILDREN IN THE UNTRANSPORTED EXPERIMENTAL GROUPS IMPROVED SIGNIFICANTLY. IN INTEGRATED NEIGHBORHOODS, RAISING THE SELF-CONCEPT LEVEL OF ONE RACE APPEARS TO RAISE THE SELF-CONCEPT LEVEL OF THE OTHER RACE AS WELL. IMPLICATIONS OF THE STUDY FOR CURRICULUM DEVELOPERS ARE DISCUSSED. STUDY LIMITATIONS ARE CONSIDERED. THIS PAPER WAS PRESENTED AT THE ANNUAL MEETING OF THE AMERICAN EDUCATIONAL RESEARCH ASSOCIATION, CHICAGO, ILLINOIS, FEBRUARY, 1968. (PS)