THIS STUDY ASSESSED VARIABLES TAKEN FROM ITEMS OF INTAKE AND DISCHARGE FORMS FOR A SAMPLE OF DISCHARGEES FROM NEW YORK STATE HOMES, CAMPS, AND START CENTERS IN 1969 AND 1970 WHO WERE DISCHARGED AFTER THE AGE OF 16. THE MAIN STATISTICAL PROCEDURE USED WAS MULTIPLE REGRESSION ANALYSIS. VARIABLES WERE ASSESSED FOR THEIR EFFECTS INDEPENDENT OF THE EFFECTS OF OTHER VARIABLES IN THE ANALYSES. THE VARIABLES REFLECTED THESE AREAS: (A) CHARACTERISTICS AS OF TIME OF REFERRAL OR ADMISSION AGE, PAST OFFENSE RECORD, LEGAL STATUS AT TIME OF REFERRAL, SCHOOL, EMPLOYMENT, FAMILY, ETHNICITY, REFERRAL COUNTY, MAIN SOURCE OF FAMILY INCOME (B) PROGRAM ACTIVITY - PROGRAM FROM WHICH YOUTH WAS DISCHARGED, DISCHARGE STATUS AND DURATION IN PROGRAM. IN AN EARLIER STUDY, SEVERAL VARIABLES PREDICTIVE OF POST-DISCHARGE ARREST AND COMMITMENT HAD BEEN DETERMINED. FOR ARREST THESE WERE NUMBER OF PREVIOUS PETITIONS, SCHOOL BEHAVIOR PROBLEMS, PRINCIPAL SOURCE OF FAMILY INCOME, CURRENT REMAND, PRESENT PETITION AND DISCHARGE STATUS. FOR COMMITMENT THESE WERE NUMBER OF PREVIOUS PETITIONS, CURRENT REMAND, PRESENT PETITION, CURRENT SCHOOL STATUS AND DISCHARGE STATUS. THE TWO SETS OF VARIABLES WERE USED AS SCALES BY DICHOTOMIZING EACH VARIABLE AND ASSIGNING ONE POINT TO THE LIMITED SOLELY TO THE REFERRAL VARIABLES. THE SCALES WERE EXAMINED WITH RESPECT TO OUTCOMES CALLED ARREST, SERIOUS ARREST, COMMITMENT AND SERIOUS COMMITMENT. THE SCALES WERE FOUND TO DIFFERENTIATE THE 1969-1970 DISCHARGEES. USING MULTIPLE REGRESSION ANALYSIS IT WAS FOUND THAT, WITH ONE POSSIBLE EXCEPTION, EACH VARIABLE CONTRIBUTED INDEPENDENTLY TO THE PREDICTIVENESS OF THE COMPOSITE SET OF VARIABLES. THE FINDINGS WERE TAKEN TO SUPPORT THE GENERAL APPROACH OF SEEKING OUT UNIQUE PREDICTORS OF ARREST AND COMMITMENT AND THEN CONSTRUCTING A SCALE COMPOSED OF THESE PREDICTORS WITH THE EXPECTATION THAT RELATIONSHIPS IN GENERAL WOULD HOLD OVER TIME. (AUTHOR ABSTRACT MODIFIED)