BROOKE HOUSE, WHICH ACCEPTS PRERELEASE STATUS PLACEMENTS FROM FEDERAL INSTITUTIONS AND STATE PRISON PAROLEES, EMPLOYS REALITY THERAPY AND EMPHASIZES JOB PLACEMENT, WORK HABITS, AND SOUND FINANCIAL PLANNING. A BASE EXPECTANCY METHOD IS USED TO EXAMINE RECIDIVISM RATES FOR 235 STATE PRISON PAROLEES WHO ENTERED THE BROOKE HOUSE PROGRAM FROM ITS INCEPTION IN 1965 THROUGH 1972. APPROXIMATELY HALF (115) OF THE CLIENTS COMPLETED THE BROOKE HOUSE PROGRAM; BUT 120 DID NOT. ROUGHLY ONE-HALF OF THE 1965-1968 CLIENTS WERE REINCARCERATED IN THE 2 YEARS DURING WHICH THEIR PROGRESS WAS MONITORED; ROUGHLY ONE-THIRD OF THE 1969-1972 CLIENTS WERE RETURNED TO PRISON DURING THE YEAR AFTER THEIR CONDITIONAL PAROLE TO BROOKE HOUSE. RECIDIVISM RATES ARE CONSIDERABLY LOWER FOR CLIENTS WHO COMPLETED THE PROGRAM (23 PERCENT FOR THE EARLIER GROUP, 14 PERCENT FOR THE LATER GROUP). RELATIVELY FEW VARIABLES DISCRIMINATE BETWEEN RECIDIVISTS AND NONRECIDIVISTS. MANY OF THE VARIABLES THAT DO RELATE TO RECIDIVISM ARE RELEVANT TO WHETHER A CLIENT COMPLETES THE PROGRAM, AND THIS, IN TURN, IS HIGHLY RELEVANT TO RECIDIVISM. COMPARISONS BETWEEN ACTUAL AND EXPECTED REIMPRISONMENT RATES FOR THE TWO GROUPS OF CLIENTS LEAD TO THE CONCLUSION THAT, AFTER ALL METHODOLOGICAL DRAWBACKS ARE CONSIDERED, THE DATA ON BROOKE HOUSE REVEAL NO NET IMPACT FOR THE PROGRAM IN TERMS OF REDUCING RECIDIVISM. IMPLICATIONS OF THE FINDINGS, DETAILS OF METHODOLOGY, AND METHODOLOGICAL PROBLEMS ARE DISCUSSED. SUPPORTING DATA ARE INCLUDED. (LKM)