MANY RESEARCHERS HAVE CONDUCTED STUDIES TO DETERMINE IF MARATHON GROUPS ARE EFFECTIVE IN PRODUCING ATTITUDINAL AND BEHAVIORAL CHANGES AMONG THE MEMBERS OF DIFFERENT POPULATIONS; THIS FORM OF TREATMENT HAS BEEN USED SUCCESSFULLY WITH BUSINESS MEN AND JUVENILE DELINQUENTS. IN THIS STUDY, THE SUBJECTS WERE 22 BLACK FEMALE HEROIN ADDICTS WHO WERE SERVING SENTENCES IN A SOUTHEASTERN CORRECTIONAL INSTITUTION IN JUNE 1976. ALL HAD BEEN IMPRISONED FOR OFFENSES RELATED TO THEIR HEROIN PROBLEMS, SUCH AS HEROIN SALES, GRAND LARCENY, ARMED ROBBERY, OR SHOPLIFTING. NONE OF THE INMATES HAD PARTICIPATED IN REGULAR COUNSELING WHILE IN PRISON. TWO GROUP LEADERS, BOTH EXPERIENCED DRUG COUNSELORS, CONDUCTED THE SESSIONS. THEY ACTED AS FACILITATORS, PLACING THE RESPONSIBILITY FOR THE SUCCESS OR FAILURE OF THE MARATHON ON GROUP PARTICIPANTS. THE GROUP APPEARED TO PROGRESS THROUGH THE FOLLOWING STAGES: RELAXATION, HOSTILITY OR PROJECTION, BEGINNING OF INTIMACY, BEGINNING OF FEEDBACK, DISCUSSION OF PROBLEMS, PROVISION OF FEEDBACK, MODIFYING BEHAVIORS, ELATION OR RELAXATION, AND ENDING STAGE. WITH REGARD TO ASSESSMENT MEASUREMENTS, A SEMANTIC DIFFERENTIAL WAS DEVELOPED TO EVALUATE THE IMPACT OF MARATHON GROUP COUNSELING IN ATTITUDES RELATED TO SELF, SIGNIFICANT OTHERS, DRUGS, THE PAST AND FUTURE, AUTHORITY, AND COUNSELING. IN ADDITION, ROTTER'S INTERNAL-EXTERNAL ORIENTATION SCALES WERE USED FOR EVALUATION OF THE SUBJECTS' ORIENTATION. POSTTESTS WERE ADMINISTERED TO BOTH THE EXPERIMENTAL AND CONTROL GROUP SUBJECTS THE DAY FOLLOWING THE MARATHON. LIMITED DIFFERENCES WERE NOTED BETWEEN THE MARATHON AND THE CONTROL GROUP WITH REGARD TO SEMANTIC DIFFERENTIAL. MARATHON GROUPS DID NOT ALTER THE ATTITUDES OF THIS POPULATION ON SUCH CONCEPTS AS AUTHORITY. HOWEVER, THE MARATHON GROUP RESPONDED MORE POSITIVELY TO THE EVALUATIVE SCALE FOR COUNSELING, AND THIS SUCCESS PROVIDES THE BASIS FOR CONCLUDING THAT MARATHON GROUPS MAY BE VALUABLE. FURTHER RESEARCH IS SUGGESTED. REFERENCES AND STATISTICAL DATA ARE PROVIDED. (LWM)