A COMPREHENSIVE PREASSESSMENT AND POSTASSESSMENT WERE CARRIED OUT ON BOTH THE SUBJECTS AND A CONTROL GROUP OF SUBJECTS ON A WAITING LIST FOR THE PROGRAM. PATIENTS AVERAGED AGE 29 AND HAD BEEN AT BROADMOOR HOSPITAL IN ENGLAND FOR AN AVERAGE OF 4.5 YEARS. BOTH GROUPS WERE ADMINISTERED A QUESTIONNAIRE PROVIDING A SCORE ON A SOCIAL ANXIETY AND DISTRESS SCALE AND A 'FEAR OF NEGATIVE EVALUATION' SCALE. NEXT, THEY LISTENED TO A SERIES OF TAPED SEQUENCES DESCRIBING A SITUATION TO WHICH THE SUBJECT RESPONDED. THIRDLY, VIDEO EQUIPMENT WAS USED TO PROMOTE CERTAIN ROLE PLAYING SITUATIONS. FINALLY, TELEPHONE SITUATIONS WERE USED TO CONTINUE THE ROLE PLAYING SCENARIOS. THE TREATMENT GROUP WAS SEEN ON A WEEKLY BASIS OVER A PERIOD OF 3 MONTHS. MUCH OF THE FORMAT INVOLVED ROLE PLAYING, FEEDBACK FROM THE VIDEOTAPES, AND REINFORCEMENT OF PROPER SOCIAL ACTIVITY. BOTH NONVERBAL AND VERBAL ASPECTS OF APPROPRIATE SOCIAL BEHAVIOR WERE TREATED. SUBJECTIVELY, CHANGES IN ALL PATIENTS WERE CLEAR AND CONVINCING. FOR OBJECTIVE MEASURES, THE FOUR TESTS WERE READMINISTERED TO THE TREATMENT GROUP AFTER TREATMENT AND THE CONTROL GROUP AFTER 3 MONTHS WITHOUT TREATMENT. POSTTREATMENT QUESTIONNAIRES SHOWED THAT THE TREATMENT GROUP REPORTED LESS ANXIETY IN SOCIAL SITUATIONS, WERE LESS CONCERNED ABOUT NEGATIVE SOCIAL EVALUATIONS, AND REPORTED LESS DIFFICULTY IN DEALING WITH A VARIETY OF SOCIAL SITUATIONS. IN THE TAPED SEQUENCES, THE TREATMENT GROUP IMPROVED ON NUMBER OF WORDS SPOKEN, LATENCY OF RESPONSE, FLUENCY ERRORS AND SOCIAL PERFORMANCE. EQUALLY ENCOURAGING RESULTS WERE SHOWN IN THE POSTTREATMENT VIDEOTAPES AND PHONE CONVERSATIONS. OVERALL, THE STRUCTURED BEHAVIORAL SKILLS PROGRAM WAS EFFECTIVE IN BRINGING ABOUT IMPROVEMENT AND IN BEING ACCEPTABLE TO THE PATIENTS. STILL QUESTIONABLE ARE GENERALIZATIONS OF THESE SKILLS TO REAL LIFE SITUATIONS AND DETERMINATIONS OF APPROPRIATE AND INAPPROPRIATE SOCIAL BEHAVIORS, GIVEN THE CHANGING TIMES AND ACCEPTABLE MANNERS. GRAPHS AND SIX REFERENCES SUPPLEMENT THE TEXT.