SEVERAL EXPERIMENTS ON THE CONCEPTUAL BEHAVIOR OF LOWER AND MIDDLE CLASS CHILDREN ARE DESCRIBED IN AN EFFORT TO CLARIFY AN APPARENT DISCREPANCY BETWEEN PREVIOUS LABORATORY FINDINGS AND OBSERVED CLASSROOM PERFORMANCE. IN THE FIRST EXPERIMENT, INDUCTIVE CONCEPT LEARNING WAS INVESTIGATED AS A FUNCTION OF SOCIAL CLASS MEMBERSHIP AND PRIOR EXPERIENCE. THE EFFECTS OF TRAINING WERE EQUAL IN THE TWO SOCIOECONOMIC GROUPS. THERE WAS NO SIGNIFICANT DIFFERENCE BETWEEN THE TWO SOCIAL CLASSES BEFORE OR AFTER TRAINING. A SECOND EXPERIMENT WAS CONDUCTED TO DETERMINE IF THE LACK OF EXPECTED SOCIAL CLASS DIFFERENCES WAS DUE TO THE METHOD OF SUBJECT SELECTION WHICH EXCLUDED CHILDREN WITH VERY HIGH OR LOW INTELLIGENCE QUOTIENTS OR TO THE FACT THAT TASK INSTRUCTION WAS ESPECIALLY EXPLICIT. DATA ANALYSIS REVEALED THAT BOTH INSTRUCTIONS AND SOCIAL CLASS CONTRIBUTED SIGNIFICANTLY TO THE TOTAL VARIANCE. SINCE TRANSFER EFFECTS ASSOCIATED WITH PRETRAINING WERE OF EQUAL MAGNITUDE FOR BOTH GROUPS, THE IDEA THAT CULTURALLY DEPRIVED CHILDREN ARE DEFICIENT IN THE CAPACITY FOR SELF-INSTRUCTION IS DISMISSED. STILLER'S FINDINGS THAT THE PERFORMANCE OF THE EXTREME ABILITY SUBJECTS ACCOUNTS FOR SOCIAL CLASS DIFFERENCES IN PERFORMANCE ARE SUPPORTED. RESULTS INDICATE THAT SOCIAL CLASS STATUS PLAYS NO PART IN CONCEPT LEARNING. THIS PAPER WAS PRESENTED AT THE AMERICAN PSYCHOLOGICAL ASSOCIATION CONVENTION, WASHINGTON, D.C., SEPTEMBER 1, 1967. (PR)