DATA WERE COLLECTED FROM A RANDOM SAMPLE OF PATIENTS TREATED IN 12 CLINICS FROM 4 COMMUNITIES (NEW ORLEANS, SAN FRANCISCO, RALEIGH/DURHAM, N.C., AND BOSTON). FOUR OF EACH OF THE FOLLOWING THREE TYPES MADE UP THE CLINIC SAMPLE: FREESTANDING DRUG CLINICS (TYPE I), DRUG UNITS IN CMHC'S (TYPE II), AND CMHC'S WITHOUT SEPARATE FACILITIES FOR DRUG ABUSERS (TYPE III). THE SAMPLE CONSISTED OF 1,113 CLIENTS INCLUDING 281 NONOPIATE ABUSERS FROM 10 OF THE 12 CLINICS; 2 OF THE TYPE III CLINICS COULD NOT IDENTIFY AN ADEQUATE SAMPLE. THE MAJORITY OF CLIENTS WERE WHITE AND UNMARRIED, AND 68 PERCENT WERE HIGH SCHOOL GRADUATES. DATA FROM A TYPE III PROGRAM IN DUMONT, N.J., WERE LATER ADDED FOR DESCRIPTIVE PURPOSES. COMPARISONS WERE MADE OF THE LARGER AND NONOPIATE SAMPLES FOR DEMOGRAPHIC CHARACTERISTICS, PROBLEMS, DRUG INVOLVEMENT, LENGTH OF TREATMENT, AND DISCHARGE DATA. DATA SOURCES WERE PATIENT RECORDS AND STAFF INTERVIEWS. INTERVIEWS INDICATED THAT STAFF OF TYPE I AND II CLINICS WERE MORE KNOWLEDGEABLE AND BETTER EXPERIENCED IN DRUG ABUSE TREATMENT, AND THAT TYPE III STAFFS HAD NO TRAINING AND MINIMAL EXPERIENCE IN THIS AREA. ONLY 1 OF THE 12 CLINICS STUDIED HAD ORGANIZED STAFF TRAINING IN IDENTIFICATION, TREATMENT, OR MANAGEMENT OF DRUG ABUSERS. AT NO CLINIC INVESTIGATED DID NONOPIATE ABUSERS RECEIVE A SPECIALIZED TREATMENT REGIMEN; INDIVIDUAL COUNSELING WAS USED MOST COMMONLY. PROGNOSIS AND PERHAPS TREATMENT APPEAR TO BE RELATED TO THE TYPE OF PROGRAM AN INDIVIDUAL ENTERS. IN ADDITION, CURRENT RESEARCH AND EVALUATION INTO THE EFFECTIVENESS OF THERAPEUTIC EFFORTS ARE PROBLEMATIC BECAUSE OF UNSTANDARDIZED RECORDKEEPING, PARTICULARLY ON DRUG USE IN THE MENTAL HEALTH CENTERS, AND ON TREATMENT PROGRESS IN ALL TYPES OF CLIENTS. TABULAR DATA ARE PROVIDED. (AOP)