SINCE THE TREATMENT OF JUVENILE OFFENDERS THROUGH INSTITUTIONALIZATION AND PROBATION HAS PRODUCED DISAPPOINTING RESULTS (RECIDIVISM RATES FOR YOUTHS EXPOSED TO THESE PROGRAMS RANGE FROM 40 PERCENT TO 55 PERCENT), OTHER TREATMENT METHODS ARE INDICATED. IN THIS MULTIPLE-BASELINE DESIGN ACROSS GROUPS STUDY, 10 YOUTHS (6 MALE, AND 4 FEMALE), RANGING IN AGE FROM 12 TO 16 YEARS, AND HAVING AN EXPERIENCE OF 1.9 PREVIOUS TREATMENT PROGRAMS, WERE GIVEN BEHAVIOR MODIFICATION, SOCIAL SKILLS, AND SELF-MANAGEMENT TRAINING. TRAINING CLASS SESSIONS MET FOR 1.5 TO 2 HOURS WEEKLY FOR 6 WEEKS. THEY CONSISTED OF WRITTEN LESSONS, DISCUSSION OF THE LESSON MATERIAL, A SHORT WRITTEN QUIZ, AND MODELING ROLE-PLAYING EXERCISES. THE FINAL SESSION CONSISTED OF A HOME VISIT IN WHICH THE YOUTHS WERE ASKED TO NEGOTIATE A CONFLICT SITUATION WITH THEIR PARENTS. MEASURES OF THE YOUTHS' BEHAVIOR INCLUDED PARENT AND TEACHER RATING SCALES, COURT RECORDS, AND SCHOOL ATTENDANCE AND GRADES. PARENTS, YOUTHS, AND THE REFERRING COMMUNITY AGENCIES ALSO FILLED OUT PROGRAM EVALUATION QUESTIONNAIRES. SINCE THE YOUTHS SUCCESSFULLY MODIFIED THEIR BEHAVIOR, DEMONSTRATING A DECREASE IN THE NUMBER OF PROBLEM BEHAVIORS EXHIBITED AT HOME AND IN SCHOOL, THE PROGRAM WAS JUDGED EFFECTIVE. INFORMATION SUGGESTS THAT THE TRAINING CLASS MAY BE A VIABLE ALTERNATIVE FOR TREATMENT. BUT FURTHER SYSTEMATIC RESEARCH IS REQUIRED TO DETERMINE ITS FULL RANGE OF APPLICABILITY. GRAPHS, TABULAR DATA, AND APPROXIMATELY 30 REFERENCES ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--MFE)