THE STUDY SOUGHT TO DELINEATE CURRENT (E.G., ADJUSTMENT IN PRISON) AS WELL AS HISTORICAL (E.G., CRIMINAL RECORD) VARIABLES PREDICTIVE OF PERFORMANCE IN A WORK RELEASE PROGRAM. THE STUDY SUBJECTS WERE MALES INCARCERATED AT THE ROBERT F. KENNEDY YOUTH CENTER AND EMPLOYED IN THE NEARBY UNIVERSITY COMMUNITY OF MORGANTOWN, W. VA. TWO SAMPLES -- A PILOT SAMPLE FOR INITIAL SELECTION OF PREDICTORS OF WORK RELEASE SUCCESS, AND A VALIDATION SAMPLE -- WERE CHOSEN. THE 190 STUDY SUBJECTS ALL HAD BEEN CONVICTED OF NONVIOLENT CRIMES (AUTO THEFT, SELLING DRUGS, BANK ROBBERY). THE BEST PREDICTORS OF SUCCESS ON WORK RELEASE WERE FOUND TO BE THE LEVEL OF INSTITUTIONAL ADJUSTMENT IMMEDIATELY PRIOR TO ENTERING THE PROGRAM, WHETHER THE INMATE HAD BEEN CONVICTED OF AUTO THEFT, HISTORY OF DRUG USE, AND AGE AT FIRST ARREST. TABULAR DATA AND A LIST OF REFERENCES ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--LKM)