IT WAS HYPOTHESIZED THAT THERE WOULD BE NO SIGNIFICANT DIFFERENCES BETWEEN THE BLACK SUBJECTS AND THE WHITE SUBJECTS AS MEASURED BY THE SEMANTIC DIFFERENTIAL TECHNIQUE AND QUESTIONNAIRE DATA AFTER A ONE-YEAR FOLLOW-UP PERIOD. THE EARLIER STUDY INVOLVED GROUP COUNSELING WITH 24 INSTITUTIONALIZED DELINQUENT FEMALES (10 BLACK AND 14 WHITE). NINETEEN SUBJECTS (NINE BLACKS, TEN WHITES) COMPLETED THE FOLLOW-UP QUESTIONNAIRES. THE POST-TEST MEANS OF THE FOUR EXPERIMENTAL CONCEPTS OF THE BLACK AND WHITE GROUPS FROM THE EARLIER STUDY WERE USED AS INDICES OF THE SUBJECT'S CONNOTATIVE MEANINGS AT THE END OF THE EXPERIMENTAL TREATMENT. COMPARISON OF THE MEAN CHANGE FROM POST-TEST TO FOLLOW-UP FOR EACH OF THE FOUR CONCEPTS REVEALED THAT THE WHITE GROUP'S MEAN SCORES WERE SIGNIFICANTLY MORE NEGATIVE AFTER ONE YEAR THAN THE BLACK GROUP'S MEAN CHANGE OF THREE OF THE EXPERIMENTAL CONCEPTS (ALL EXCEPT 'FATHER'). THE COMPARATIVE BEHAVIOR OF THE BLACK AND WHITE SUBJECTS DURING THE YEAR'S FOLLOW-UP REFLECTED THE RESULTS OF THE SEMANTIC DIFFERENTIAL. THE BLACK GROUP WAS SUPERIOR IN RELATION TO THE NUMBER OF INSTITUTIONAL RELEASES AND NON-VIOLATION OF PROBATION/PAROLE. IT IS CONCLUDED THAT, CONTRARY TO THE RESULTS OF PREVIOUS STUDIES, THE LONG-TERM EFFECTS OF GROUP COUNSELING WERE SUBSTANTIALLY SUPERIOR FOR THE BLACK FEMALES. A LIST OF REFERENCES IS INCLUDED. (AUTHOR ABSTRACT MODIFIED)