THE PRINCIPAL OBJECTIVES OF THIS STUDY WERE (1) TO DESCRIBE DEVELOPMENT STAGES OF STUDENTS FROM FIRST- THROUGH NINTH-GRADE LEVELS AS THEY AFFECT THE SOLUTION OF A THREE-DIMENSIONAL CLASSIFICATION PROBLEM, (2) TO UTILIZE THE INHELDER-PIAGET AND THE BRUNNER SYSTEMS OF DESCRIBING STAGES OF COGNITIVE DEVELOPMENT, AND (3) TO DETERMINE BY MULTIVARIATE ANALYSIS IF DIFFERENCES IN DEVELOPMENTAL STAGES ARE RELATED TO VERBAL AND NONVERBAL OR PERFORMANCE APTITUDES. SUBJECTS OF GRADE LEVELS 1 THROUGH 9 (360) SORTED 54 OBJECTS (VARIED SIZES, COLORS AND SHAPES) INTO A 3-DIMENSIONAL MATRIX, WITH MOVEABLE BOXES OF 3 LEVELS, 3 ROWS AND 6 COLUMNS. AN ADDITIONAL GROUP OF SUBJECTS RECEIVED DISCRIMINATION TRAINING ON THE OBJECTS. THE RESULTS SHOWED THAT SIGNIFICANT AGE-RELATED INCREASES IN SCORES OCCURRED, FIRST FOR COLOR, THEN SIZE, AND THEN FOR FORM. HORIZONTAL SCORES PRECEDED VERTICAL. THERE WERE NO SEX DIFFERENCES IN ANY OF THE GRADES OR TASKS. THE USE OF LARGE OBJECTS IN EARLY SORTING GAVE EVIDENCE OF "CONCRETE" RESPONSES EXTENDING INTO THE NINTH GRADE. "SCANNING" STRATEGY WAS LATE IN DEVELOPING. DISCRIMINATION TRAINING SIGNIFICANTLY INCREASED THE SCORES. (JC)