RELATIONSHIPS BETWEEN MENTAL AGE LEVEL AND DEGREE OF SOCIAL PARTICIPATION IN TRAINABLE AND EDUCABLE MENTALLY RETARDED CHILDREN OF BOTH SEXES WERE INVESTIGATED. THE MAJOR OBJECTIVES WERE (1) TO ESTABLISH A GRADIENT OF SOCIAL PARTICIPATION IN MENTALLY RETARDED CHILDREN OF BOTH SEXES REPRESENTING PUBLIC SCHOOL AND INSTITUTIONAL ENVIRONMENT, (2) TO COMPARE THE SOCIAL PARTICIPATION OF TRAINABLE AND EDUCABLE GROUPS TO NORMS ESTABLISHED FOR CHILDREN OF NORMAL INTELLIGENCE, AND (3) TO RELATE DEGREES OF SOCIAL PARTICIPATION WITH THE EDUCATIONAL ENVIRONMENTS OF EDUCABLE AND TRAINABLE MENTALLY RETARDED CHILDREN. FOUR GROUPS OF MENTALLY RETARDED CHILDREN (120) WERE SELECTED FROM PUBLIC SCHOOL AND INSTITUTIONAL CLASSES. TWO GROUPS (TRAINABLE AND EDUCABLE) REPRESENTED PUBLIC SCHOOL CLASSES AND INSTITUTIONAL CLASSES. DURING THE 5-MONTH PILOT STUDY ALL GROUPS WERE OBSERVED AND THEIR BEHAVIOR WAS NOTED FOR A 1-MINUTE FREE PLAY PERIOD. THESE OBSERVATIONS WERE GATHERED AT WEEKLY AND SEMIWEEKLY INTERVALS. TRAINED OBSERVERS RATED THE BEHAVIOR OF THE SUBJECTS IN SUCH CATEGORIES AS UNOCCUPIED, SOLITARY, ONLOOKER, ASSOCIATIVE, OR COOPERATIVE. ALL GROUPS WERE AGAIN RATED AT THE END OF THE SCHOOL YEAR. AN ANALYSIS WAS MADE OF THE CHANGE IN RATINGS FOR ALL GROUPS FROM INITIAL TO FINAL SCORES. THE RESULTS OF THE FACTORIAL DESIGN DEMONSTRATED A STATISTICALLY SIGNIFICANT DIFFERENCE IN FAVOR OF EDUCABLE CHILDREN OVER TRAINABLE CHILDREN IN TOTAL PLAY BEHAVIOR SCORES. (GC)