THE FIRST HYPOTHESIS OF THE STUDY ASSUMED THAT IF THREE COMPARABLE GROUPS OF PAROLE ADDICTS WERE EXPOSED TO DIFFERENTIAL COMMUNICATION SITUATIONS VARYING ALONG RELEVANT DIMENSIONS, ONE COULD EXPECT PREDICTABLE BEHAVIORAL RESPONSES QUANTIFIABLE IN TERMS OF PAROLE VIOLATION RATES. A PAROLED ADDICT WAS DEFINED IN TERMS OF PRIMARY DRUG ADDICTION, WHICH REFERS TO A COMPLEX BEHAVIOR PATTERN ASSOCIATED WITH THE COMPULSIVE USE OF AN OPIATE. MOST ADDICTS IN THE SAMPLE WERE OF MEXICAN DESCENT, THE MEAN AGE BEING 30 YEARS. THREE TREATMENT MODALITIES WERE ESTABLISHED: INDIVIDUAL COUNSELING APPROACH, FIELD EXAMINATION APPROACH, AND ROUTINE SUPERVISION APPROACH. THE SECOND HYPOTHESIS OF THE STUDY, FORMULATED ABOUT 11 YEARS AFTER THE TERMINATION OF THE EXPERIMENTAL PERIOD, WAS BASED ON THE GENERALLY PESSIMISTIC VIEW OF THE POSSIBILITY OF LONG-RANGE REHABILITATION FOR ADDICTS. ADDICTS IN THE INDIVIDUAL COUNSELING GROUP WERE SCHEDULED FOR REGULAR 1-HOUR INTERVIEWS WEEKLY OR EVERY 2 WEEKS. ADDICTS IN THE FIELD EXAMINATION GROUP WERE TOLD TO EXPECT AT LEAST TWO SURPRISE HOME VISITS PER MONTH. ADDICTS IN THE ROUTINE SUPERVISION OR CONTROL GROUP WERE SEEN SOMEWHAT LESS THAN ADDICTS IN THE FIELD EXAMINATION GROUP, AND NO ROUTINE PHYSICAL EXAMINATIONS WERE PERFORMED. RESPONSE TO TREATMENT FOR THE EXPERIMENTAL STAGE WAS MEASURED BY PAROLE VIOLATION RATES IN TERMS OF A SPECIALLY CONSTRUCTED PAROLE VIOLATION INVENTORY. THE DEPENDENT VARIABLE FOR THE FOLLOWUP STAGE 11 YEARS AFTER THE EXPERIMENTAL STAGE WAS DEFINED BY THE EXTENT OF POSTEXPERIMENTAL ADJUSTMENT OF ADDICTS, AS EVIDENCED BY CRIMINAL INVOLVEMENT RATES. FROM THE PERSPECTIVE OF CONTRACULTURAL ADDICTION AS A BEHAVIOR PATTERN, INFERENCES FROM THE DATA AND REVIEW OF THE LITERATURE INDICATED THAT CORRECTIONAL MANAGEMENT AND TREATMENT OF CONTRACULTURAL ADDICTS IS FEASIBLE BUT LIMITED TO THE ACTUAL EXPOSURE PERIOD AND THAT A DRUG-FREE CORRECTIONAL PROGRAM HIGHLY EFFECTIVE IN TERMS OF SHORT-TERM RESULTS IS MORE LIKELY TO FAIL IN THE LONG RUN. SUPPORTING DATA AND REFERENCES ARE INCLUDED. (DP)