DARP WAS SET UP AS A PATIENT REPORTING AND TRACKING SYSTEM TO EVALUATE DRUG TREATMENT GROUPS. THE PATIENTS WERE ENROLLED IN ONE OF FIVE TREATMENT REGIMES--OUTPATIENT METHADONE MAINTENANCE (MM), RESIDENTIAL THERAPEUTIC COMMUNITY (TC), OUTPATIENT DRUG-FREE TREATMENT (DF), OUTPATIENT DETOXIFICATION (DT), AND INTAKE ONLY (IO). THE IO GROUP SERVED AS A COMPARISON BUT WAS NOT A TRUE CONTROL GROUP BECAUSE THE MEMBERS WERE NOT RANDOMLY ASSIGNED. STATISTICALLY SIGNIFICANT IMPROVEMENT IN DRUG USE, EMPLOYMENT, AND CRIMINALITY INDICATORS WERE FOUND FOR THE MM, TC, AND DF GROUPS BUT NOT THE DT AND IO. OPIOID USE DECREASED IN MM, TC, AND DF. HOWEVER, MARIJUANA USE INCREASED IN MM. ALCOHOL USE INCREASED IN ALL EXCEPT DF, ALTHOUGH NOT TO PROBLEM DIMENSION. EMPLOYMENT INCREASED IN MM, TC, AND DF AND THE ARREST RATE DECREASED IN MM. BEFORE ENTERING THE PROGRAMS 60 PERCENT HAD RECEIVED NO TREATMENT; IN THE FIRST 3 YEARS OF FOLLOWUP, 51 PERCENT REPORTED AT LEAST ONE REENTRY TO DRUG TREATMENT. LOW CRIMINAL BACKGROUND FACTORS, LOW SOCIAL DEVIANCE, LONGER TENURES IN TREATMENT, AND STABLE FAMILY BACKGROUND WERE ASSOCIATED WITH FAVORABLE OUTCOMES, WHILE HIGH CRIMINAL BACKGROUNDS WERE ASSOCIATED WITH LEAST FAVORABLE OUTCOMES. CHARTS ILLUSTRATE THESE ANALYSES. REFERENCES ARE APPENDED. (FOR A MORE DETAILED DESCRIPTION OF DARP, SEE NCJ-60590). (GLR)