FOR COMPARATIVE PURPOSES, THE 33 RESIDENTS RELEASED IN 1977 AND THE 55 INDIVIDUALS RELEASED IN 1978 WERE DIVIDED INTO PROGRAM COMPLETERS AND NONCOMPLETERS. THE FIRST GROUP REFERS TO THOSE WHO SUCCESSFULLY COMPLETED THEIR STAY AT THE DRUG HOUSE AND WERE RELEASED EITHER BY PAROLE BOARD PERMIT OR BY A CERTIFICATE OF DISCHARGE. THE LATTER GROUP CONSISTED OF THOSE WHO DID NOT COMPLETE THEIR STAY BECAUSE OF RULES VIOLATION, ADJUSTMENT PROBLEMS, ESCAPE ATTEMPTS, OR DISCIPLINARY PROBLEMS. OF THE 1977 RELEASEE SAMPLE, 15 (45 PERCENT) WERE SUCCESSFUL PROGRAM COMPLETERS, WHILE FOR THE 1978 SAMPLE 27 (49 PERCENT) WERE SUCCESSFUL COMPLETERS. DATA WERE COLLECTED FROM SUBJECTS' CRIMINAL HISTORIES, SOCIAL BACKGROUNDS, AND COMMITMENT VARIABLES. MULTIVARIATE ANALYSIS USING THESE VARIABLES WAS CARRIED OUT. CHARACTERISTICS OF INDIVIDUALS WITH A HIGH PROBABILITY OF SUCCESS TENDED TO BE THOSE WHO WERE FIRST ARRESTED AS ADULTS WITH NO PRIOR PROPERTY OFFENSE, WERE MARRIED, HAD TWO OR MORE FURLOUGHS AND SUCCESSFUL FURLOUGH OUTCOMES, AND WERE TRANSFERRED FROM A MEDIUM OR MINIMUM SECURITY FACILITY. THOSE WITH A MODERATE PROBABILITY OF SUCCESS WERE AGE 30 OR OLDER AT RELEASE DATE AND AGE 25 OR OLDER AT INCARCERATION, HAD SPENT 5 MONTHS OR MORE AT THEIR LONGEST AND AT THEIR MOST SKILLED POSITION, HAD COMPLETED 10TH GRADE OR MORE, HAD 11 OR FEWER PRIOR COURT APPEARANCES AND 3 OR FEWER PERSON OFFENSES, SERVED A YEAR OR LESS BEFORE BEING PLACED IN A PRERELEASE FACILITY. THE REMAINING INDIVIDUALS HAD A LOW PROBABILITY OF SUCCESS IN A DRUG CONTRACT HOUSE. EXTENSIVE TABLES ARE PROVIDED.