DATA WERE GATHERED ON A RANDOM SAMPLE OF 250 ADULT MALE FELONS WHO HAD PARTICIPATED IN THE WORK RELEASE PROGRAM BETWEEN 1973 AND 1974. DISCRIMINANT ANALYSIS WAS USED TO DEVELOP A PREDICTIVE INSTRUMENT WHICH DIFFERENTIATES BETWEEN SUCCESSFUL AND UNSUCCESSFUL PARTICIPANTS. AFTER PRETESTING, THE INSTRUMENT WAS IMPLEMENTED IN FEBRUARY 1976 AS AN AID IN THE WORK RELEASE PROGRAM'S PARTICIPANT SELECTION PROCESS. WITHIN 1 YEAR AFTER THE INSTRUMENT'S IMPLEMENTATION, THE INPROGRAM VIOLATION RATE WAS REDUCED FROM 30 PERCENT TO APPROXIMATELY 11 PERCENT. PRIOR TO IMPLEMENTATION, 3 OUT OF EVERY 10 OFFENDERS ASSIGNED TO WORK RELEASE FAILED. OF THE 104 OFFENDERS ASSIGNED TO THE PROGRAM AFTER THE OBJECTIVE SELECTION PROCEDURE WAS IMPLEMENTED, 84 WERE PAROLED, 8 WERE DISCHARGED, 8 WERE TERMINATED FROM THE PROGRAM AS A RESULT OF PROGRAM STANDARD VIOLATIONS, 2 WERE TERMINATED AS A RESULT OF INSTITUTIONAL CHARGES OR RULE INFRACTIONS, AND 2 ESCAPED. IT IS CONCLUDED THAT THE OBJECTIVE SELECTION PROCEDURE HAS BEEN SUCCESSFUL IN IMPROVING THE WORK RELEASE PROGRAM'S OPERATIONAL EFFECTIVENESS. THE SELECTION PROCEDURE MAKES IT POSSIBLE FOR THE PROGRAM TO BE MORE CLOSELY MONITORED AND CONTROLLED. BY QUANTIFYING INFORMATION ON OPERATIONAL EFFECTIVENESS, OBJECTIVE SELECTION ENHANCES EVALUATION OF TREATMENT EFFECTIVENESS. A TABLE SHOWING VARIABLES INCLUDED IN THE SELECTION INSTRUMENT, SUPPORTING DATA, AND A LIST OF REFERENCES ARE INCLUDED. (LKM)