THREE STORIES DEALT WITH THEFT AND THREE WITH DESTROYING A BOY SCOUT CLUBHOUSE. EACH STORY OF THE GROUP OF THREE DIFFERED IN THE AMOUNT OF GROUP PRESSURE EXERTED ON THE 'HERO'. SUBJECTS WERE FIRST ASKED TO DECIDE HOW THE HERO WILL BEHAVE IN THE SITUATION, AND THEN TO INDICATE WHAT THEY THOUGHT THEIR FRIENDS HAD DECIDED. SUBJECTS WERE 214 MALE JUVENILE DELINQUENTS, COMMITTED TO EIGHT CORRECTIONAL INSTITUTIONS (AVERAGE AGE 14.3). NINETY NON-DELINQUENTS ATTENDING RESIDENTIAL INSTITUTIONS SERVED AS A CONTROL GROUP (AVERAGE AGE 14.4). RESULTS SHOW THAT DELINQUENTS GAVE MORE 'DELINQUENT' ANSWERS TO THE QUESTIONNAIRE THAN DID NON-DELINQUENTS AND THAT THEY GAVE MORE ANSWERS WHICH ASSERTED THAT THEIR FRIENDS WERE MORE COMMITTED TO DELINQUENCY THAN THEY THEMSELVES WERE. DELINQUENTS AND NON-DELINQUENTS ALIKE TOOK GROUP PRESSURE INTO ACCOUNT AS A POTENTIAL SOURCE OF INFLUENCE ON THE HEROES OF THE STORIES. THESE FINDINGS ARE DISCUSSED IN RELATION TO COHEN'S AND MATZA'S DIVERGENT THEORIES ABOUT DELINQUENT SUBCULTURE. WHILE THE FINDINGS RAISE SOME DOUBTS ABOUT THE EXISTENCE OF A DELINQUENT SUBCULTURE, THEY SUPPORT MATZA'S NOTION OF A PROCESS OF PLURALISTIC IGNORANCE PREVAILING IN WHAT HE CALLS THE SITUATION OF COMPANY. THE POSSIBLE PSYCHOLOGICAL FUNCTIONS OF DELINQUENTS MAINTAINING A DIFFERENCE BETWEEN THEMSELVES AND THEIR FRIENDS AND THE IMPLICATIONS FOR THERAPY AND REHABILITATION ARE CONSIDERED. (AUTHOR ABSTRACT)