WITHIN THE TREATMENT PROGRAM, TWO TREATMENT METHODS WERE STUDIED: TRADITIONAL SOCIAL-GROUP WORK AND GROUP-LEVEL BEHAVIORAL MODIFICATION. A THIRD SECTION OF GROUPS RECEIVED NO SYSTEMATIC TREATMENT. MOST OF THE REFERRED BOYS WERE ASSIGNED TO TREATMENT GROUPS COMPOSED SOLELY OF OTHER REFERRED ANTISOCIAL BOYS, BUT ABOUT 10 PERCENT WERE ASSIGNED TO REGULAR PROSOCIAL YOUTH GROUPS AT THE CENTER. A MAJOR OBJECTIVE OF THE PROGRAM WAS TO ASCERTAIN THE EXTENT TO WHICH TYPICAL COMMUNITY AGENCIES CAN DELIVER EFFECTIVE GROUP TREATMENT SERVICES WITHOUT, AT THE SAME TIME, MAKING MAJOR CHANGES IN THEIR USUAL STAFFING OR OPERATING PATTERNS. ANOTHER IMPORTANT GOAL WAS TO EXAMINE THE ACTUAL EFFECTS OF SUCH GROUP TREATMENT ON BOTH THE REFERRED AND THE UNREFERRED BOYS WHO TOOK PART IN THEM. IN GENERAL, THE REFERRED BOYS, THEIR PARENTS, AND THE REFERRING AGENTS REPORTED THAT MOST OF THE BOYS IMPROVED CONSIDERABLY AS A RESULT OF THE PROGRAM. YET, WEEKLY BEHAVIORAL OBSERVATIONS OF THE BOYS REVEALED RELATIVELY LITTLE BEHAVIORAL CHANGE IN ONE DIRECTION OR THE OTHER AND FEW VARIATIONS ACCORDING TO TYPE OF TREATMENT METHOD OR LEADER TRAINING. MORE IMPORTANT, PERHAPS, THE WEEKLY BEHAVIORAL OBSERVATIONS DID SHOW THAT REFERRED BOYS WHO WERE TREATED AMONG PROSOCIAL PEERS SHOWED CONSIDERABLY LESS ANTISOCIAL BEHAVIOR THAN DID THEIR COUNTERPARTS WHO WERE TREATED AMONG ANTISOCIAL PEERS. THE FORMER BOYS IN FACT SHOWED EVEN LOWER INCIDENCES OF ANTISOCIAL BEHAVIOR THAN DID THE NONREFERRED BOYS. IN CONTRAST, REFERRED BOYS WHO WERE TREATED AMONG ANTISOCIAL PEERS CONSISTENTLY DISPLAYED MORE ANTISOCIAL BEHAVIOR THAN DID ANY OF THE OTHER CATEGORIES OF CHILDREN. (AUTHOR ABSTRACT MODIFIED)