ONLY THOSE WITH A FOLLOWUP PERIOD OF AT LEAST 6 MONTHS WAS CONSIDERED. DATA WERE COLLECTED CONERNING SEX, RACE, AGE, NUMBER OF PRIOR OFFENSES, PUBLIC ASSISTANCE, AND REASON REFERRED. RECIDIVISM WAS MEASURED ON THE BASIS OF REFERRALS, ARRESTS, SEVERITY OF OFFENSE, AND WHETHER INSTITUTIONALIZATION WAS REQUIRED. RESULTS SHOWED THAT ALL TREATMENT PROGRAMS WERE MORE SUCCESSFUL FOR FEMALES THAN FOR MALES, THAT FEMALES WERE MORE LIKELY TO HAVE BEEN COMMITTED FOR STATUS OFFENSES THAN WERE MALES, AND THAT FEMALES HAD A LOWER MEAN NUMBER OF PRIOR REFERRALS THAN MALES FOR ALL TREATMENT PROGRAMS. FOR THE THREE SAMPLE PERIODS OF PROBATION EXAMINED, MALES HAVE SHOWN TRENDS OF DECREASING MEAN NUMBER OF PRIOR REFERRALS AND INCEASING MEAN AGE. PROBATION--WITH THE LOWEST MEAN NUMBER OF PRIOR REFERRALS, THE LOWEST PERCENT OF PUBLIC ASSISTANCE, AND THE SECOND HIGHEST MEAN AGE FOR MALES OF ALL THE TREATMENT PROGRAMS--HAD THE HIGHEST SUCCESS RATE FOR MALES. MALES COMMITTED TO GROUP HOMES HAD THE LOWEST MEAN AGE FOR THE MALE SAMPLES, WHILE FEMALES IN GROUP HOMES HAD THE SECOND-HIGHEST MEAN AGE. THE 2-DAY TREATMENT PROGRAMS HAD THE HIGHEST FAILURE RATES FOR MALES; MALES COMMITTED TO THE DEPARTMENT FOR HUMAN SERVICES HAD THE HIGHEST MEAN NUMBER OF PRIOR REFERRALS; AND MALES COMMITTED TO THE RESIDENTIAL PROGRAM AT ORMSBY VILLAGE (RESIDENTIAL CENTER WITH COTTAGE GROUPS) RANKED MEDIUM FOR MEAN AGE, MEAN NUMBER OF PRIOR REFERRALS, AND COMMITMENTS FOR MAJOR OFFENSES, WHILE HAVING ONE OF THE LOWEST SUCCESS RATES. TABULAR AND GRAPHIC DATA ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--RCB)