A MAIN PROBLEM ARISING IN THE STUDY OF CATEGORIZING BEHAVIOR IN PRESCHOOL CHILDREN BY MEANS OF SORTING PROCEDURES WAS THAT PERFORMANCE APPEARED TO BE VERY MUCH A FUNCTION OF THE CHILD'S COMPREHENSION OF THE PARTICULAR VERBAL INSTRUCTIONS EMPLOYED BY THE EXPERIMENTER IN PRESENTING THE TASK. THE SPECIFIC AIMS OF THE FIRST STUDY WERE TO INVESTIGATE THE NATURE AND FREQUENCY OF OCCURRENCE OF COGNITIVELY STRUCTURED OBJECT MANIPULATIONS IN INFANTS AND TO ASSESS THE ROLE OF SEVERAL DIFFERENT STIMULUS CONTRASTS IN THE INSTIGATION OF SUCH BEHAVIOR. NO SPECIFIC SORTING INSTRUCTIONS WERE GIVEN, BUT INFANTS WERE SIMPLY ENCOURAGED TO "PLAY WITH" OR "FIX" THE TEST OBJECTS. THE PURPOSES OF THE FOLLOWUP STUDY WERE TO EXAMINE CHANGES IN THE NATURE AND FREQUENCY OF OCCURRENCE OF THE SERIAL ORDERING AND OBJECT GROUPING BEHAVIORS, TO EXAMINE CHANGES IN SOME OF THE QUALITATIVE FEATURES CHARACTERIZING THE CHILDREN'S OBJECT ARRANGEMENTS, AND TO SEE WHETHER INDIVIDUAL DIFFERENCES SHOWED ANY STABILITY OVER THE DEVELOPMENTAL INTERVAL BETWEEN THE FIRST AND SECOND TESTING. THE TESTING PROCEDURE FOR EACH TASK INVOLVED PLACING THE ARRAY OF TEST OBJECTS IN FRONT OF THE CHILD, WHO WAS THEN ENCOURAGED TO PLAY WITH THEM WITH SUCH COMMENTS AS "SEE THESE, YOU PLAY WITH THEM, YOU FIX THEM ALL UP." SINCE THE PROCEDURE WAS INTENDED TO ELICIT SPONTANEOUS BEHAVIOR, THE EXAMINER INTRUDED AS LITTLE AS POSSIBLE, EXCEPT TO ENCOURAGE THE CHILD TO CONTINUE TO PLAY WITH THE OBJECTS IF HE HAD BECOME DISINTERESTED. RESULTS SHOWED THAT, AT 20 MONTHS, THE MOST COMMON ORDERING BEHAVIOR CONSISTED OF SUCCESSIVE MANIPULATION OF FOUR OBJECTS OF ONE KIND. WHEN THE INFANTS WERE RETESTED AT 40 MONTHS, SUCCESSIVE MANIPULATION OF ALL EIGHT OBJECTS IN BOTH GROUPS OCCURRED MORE THAN THREE TIMES AS FREQUENTLY AS AT 20 MONTHS, EVEN WITH THE GENERAL INSTRUCTIONS SIMPLY TO "PLAY WITH," OR "FIX" THE OBJECTS.