A MINNESOTA STUDY OF TWO JUVENILE INSTITUTIONS AND SIX RESIDENTIAL CENTERS FOUND THAT BOTH RESULTED IN REDUCED OFFENSES BUT THAT THE SHORT-TERM VERSUS LONG-TERM COSTS WERE SIGNIFICANTLY DIFFERENT. AFTER 1 YEAR, THE COMMUNITY-BASED TREATMENT CENTERS HAD A 0.6 REDUCTION IN OFFENSES FOR A COST PER REDUCED OFFENSE OF $5,463. INSTITUTIONS HAD A REDUCTION OF 2.6 FOR A COST PER REDUCED OFFENSE OF $3,832. AFTER 3 YEARS, HOWEVER, TREATMENT CENTERS HAD A REDUCED OFFENSE RATE OF 6.6 FOR A COST OF $497, WHILE THE INSTITUTIONS HAD A 9.6 REDUCTION AT A COST OF $1038. THE RECIDIVISM RATE WAS FIGURED AS A COMPOSITE OF SEVERAL WEIGHTED FACTORS INCLUDING PROBABILITY OF OFFENSES REOCCURRING AND HISTORY OF THE OFFENDER. THE STUDY ALSO FOUND THAT COSTS COULD BE REDUCED IF FACILITIES WERE OPERATED AT 90 PERCENT OF CAPACITY. AT TIME OF STUDY, THE TREAMENT CENTERS WERE OPERATING AT 83 PERCENT AND THE INSTITUTIONS AT 75 PERCENT CAPACITY. ALSO, COSTS COULD BE REDUCED WITHOUT AFFECTING RECIDIVISM BY REDUCING STAY FROM 190 DAYS TO 91 DAYS; BOTH STAYS HAD SIMILAR RATES FOR THE 3-YEAR PERIOD. WAYS OF USING SIMILAR MODELS FOR STUDIES IN OTHER COMMUNITIES ARE DISCUSSED.