PROGRAMS FOR THE TREATMENT OF HEROIN AND MARIHUANA USERS OR USERS OF OTHER DRUGS (NOT INCLUDING ALCOHOL), WHICH WERE FUNDED BY THE NATIONAL INSTITUTE ON DRUG ABUSE, THE CALIFORNIA SHORT-DOYLE PROGRAM, OR THE COUNTY WERE INCLUDED IN THIS STUDY. PROGRAMS OPERATED DIRECTLY BY THE STATE OF CALIFORNIA OR THE FEDERAL GOVERNMENT, OR FUNDED BY THE BUREAU OF PRISONS WERE EXCLUDED. THE CLIENTS WERE TO BE INTERVIEWED APPROXIMATELY 1 YEAR AFTER DISCHARGE FROM TREATMENT; QUESTIONS REGARDED DRUG USE, EMPLOYMENT, CRIMINAL INVOLVEMENT, AND PSYCHO-PHYSIOLOGICAL HEALTH. THE ANALYSIS WAS TO BE KEYED TO TREATMENT MODALITY, INCLUDING OUTPATIENT DRUG-FREE, RESIDENTIAL DRUG-FREE, OUTPATIENT METHADONE MAINTENANCE, AND INPATIENT DETOXIFICATION TREATMENT. THE DISCHARGE CRITERION WAS WAIVED FOR OUTPATIENT METHADONE MAINTENANCE. FURTHERMORE, A COST-BENEFIT ANALYSIS WAS TO BE CONDUCTED OF THE DIFFERENT KINDS OF TREATMENT, AND THOSE IN THE TREATED SAMPLE WERE TO BE COMPARED WITH THOSE GIVEN A JAIL SENTENCE RESULTING FROM DRUG USE. THE STUDY POPULATION CONSISTED OF THOSE CLIENTS ADMITTED TO TREATMENT DURING THE PERIOD FROM MARCH 1, 1975, THROUGH JULY 31, 1975, WHO HAD BEEN DISCHARGED FOR ANY REASON BY DECEMBER 31, 1975. DATA WERE COLLECTED BY INTERVIEWING CLIENTS, WITH INTERVIEWS LASTING ABOUT 1.5 HOURS. ONLY ONE-THIRD OF THE FORMER CLIENTS INCLUDED IN THE STUDY POPULATION WERE LOCATED BY THE TREATMENT AGENCIES. MOST ATTEMPTS WERE MADE BY MAIL OR TELEPHONE. EFFECTIVE CLIENT TRACKING WAS GREATLY HINDERED BY FEDERAL AND STATE LAWS AND REGULATIONS CONCERNING CLIENTS' RIGHTS TO PRIVACY. IT WAS PRESUMED THAT THE TREATMENT AGENCIES WOULD BE ABLE TO LOCATE A HIGH PROPORTION OF THEIR FORMER CLIENTS, AND ON THIS BASIS, A SAMPLE OF ABOUT 1,800 CASES WAS SELECTED, WITH THE OBJECTIVE OF COMPLETING 1,200 CASES. HOWEVER, THE ACTUAL LOCATION RATE WAS VERY LOW, RESULTING IN A NONREPRESENTATIVE SAMPLE. THUS, THE FINDINGS AND CONCLUSIONS FROM THIS STUDY CANNOT BE GENERALIZED TO THE TOTAL STUDY POPULATION. FINDINGS, CONCLUSIONS, AND RECOMMENDATIONS FOR EACH TREATMENT MODALITY ARE DISCUSSED. THE MERITS OF CONVENTIONAL TREATMENT IMPACT STUDIES (SUCH AS THIS ONE) SHOULD BE CRITICALLY EXAMINED, AND AN INTENSIVE EFFORT SHOULD BE MADE TO EXPLICATE THE EVALUATION OF DRUG-ABUSE TREATMENT FROM A DEVELOPMENT PERSPECTIVE. FIGURES AND TABLES ARE INCLUDED. (PRG)