THE PROGRAM WAS CONDUCTED IN A LABORATORY SCHOOL, WHERE STUDENTS WERE INSTRUCTED OVER A 2-YEAR PERIOD IN AN EXPERIMENTAL ENVIRONMENT, SEPARATED FROM THEIR REGULAR PUBLIC SCHOOLS. PROCEDURES BASED ON SOCIAL LEARNING THEORY WERE USED TO IMPROVE THE STUDENTS' BASIC READING, ARITHMETIC, AND INTERPERSONAL SKILLS. OF 24 YOUTHS PARTICIPATING IN THE PROGRAM, 15 COMPLETED A FOLLOWUP EVALUATION CONDUCTED APPROXIMATELY 5 YEARS LATER. THEIR ATTITUDES AND PERFORMANCE WERE ASSESSED THROUGH SELF-REPORT MEASURES, INCLUDING EMPLOYMENT AND EDUCATIONAL STATUS, EVALUATION OF PROGRAM PARTICIPATION, INVOLVEMENT IN LEISURE-TIME AND COMMUNITY ACTIVITIES, RELATIONSHIPS WITH FAMILY AND FRIENDS, AND ANTICIPATED CONSEQUENCES OF ENGAGING IN DELINQUENT ACTS. THE PARTICIPANTS EVALUATED THE PROGRAM FAVORABLY AND TENDED TO FEEL RELATIVELY HAPPY ABOUT THEIR CURRENT SITUATIONS. ABOUT TWO-THIRDS REPORTED CLOSE RELATIONSHIPS WITH FAMILY AND PEERS. HOWEVER, SECURING AND KEEPING JOBS APPARENTLY PRESENTS A CONTINUING PROBLEM. SEVERAL PARTICIPANTS ALSO REPORTED INVOLVEMENT IN EITHER STATUS OR CRIMINAL OFFENSES. OVERALL, THE DATA DO NOT DEMONSTRATE CONCLUSIVELY THE LONG-TERM MERITS OF THE PROGRAM. METHODOLOGICAL PROBLEMS IN CONDUCTING THE FOLLOWUP ARE IDENTIFIED, AND IMPLICATIONS FOR FUTURE LONG-TERM EVALUATIONS ARE DISCUSSED. A LIST OF REFERENCES IS INCLUDED. (LKM)