SINCE IT WAS THOUGHT THAT PATTERNS OF SOCIAL AND CRIMINAL BEHAVIOR MIGHT DIFFER BETWEEN MEN AND WOMEN, DATA COLLECTION AND ANALYSIS NECESSARY TO DESIGN A POINT SYSTEM WERE CARRIED OUT SEPARATELY BY SEX. THE SAMPLE OF 146 MEN INCLUDED ALL RESIDENTS AT A WORK RELEASE CORRECTIONAL FACILITY WHO WERE TEMPORARILY RELEASED DURING THE PERIOD FROM AUGUST 1974 THROUGH AUGUST 1975. THE SAMPLE OF 213 WOMEN WAS TAKEN FROM A LIST OF WORK RELEASE AND EDUCATIONAL LEAVE APPLICANTS AT ANOTHER CORRECTIONAL FACILITY FOR PARTICIPATION IN TEMPORARY RELEASE BEGINNING IN NOVEMBER 1971. TO ALLOW FOR SUFFICIENT PAROLE FOLLOWUP, THE SAMPLE INCLUDED ONLY WOMEN WHO HAD COMPLETED OR HAD BEEN REMOVED FROM THE FACILITY BY JULY 1975. TWO TYPES OF DATA WERE COLLECTED FOR EACH INDIVIDUAL IN THE SAMPLES: INFORMATION KNOWN AT THE TIME OF APPLICATION FOR TEMPORARY RELEASE THAT MIGHT BE USEFUL IN PREDICTING PERFORMANCE IN THE PROGRAM, AND MEASURES OF SUCCESS OR FAILURE IN TEMPORARY RELEASE. PREDICTOR DATA INCLUDED DEMOGRAPHIC VARIABLES, CRIMINAL HISTORY, AND BEHAVIOR IN CORRECTIONAL FACILITIES PRIOR TO PARTICIPATION IN TEMPORARY RELEASE. WHILE THE ABILITY TO PREDICT SUCCESS IN THE TEMPORARY RELEASE PROGRAM, INCLUDING MINIMIZING RISK TO THE COMMUNITY, WAS THE MOST IMPORTANT CRITERION IN DESIGNING THE POINT SYSTEM, OTHER CONSIDERATIONS WERE TAKEN INTO ACCOUNT. AFTER EVALUATING THREE ALTERNATIVE APPROACHES, A POINT SYSTEM WAS DESIGNED SO THAT TEMPORARY RELEASE SELECTION WOULD BE FAIR TO INMATES AND AS EASY AS POSSIBLE TO ADMINISTER. CARE WAS TAKEN TO MAKE USE OF INTUITIVE JUDGMENTS OF INDIVIDUALS ACQUAINTED WITH TEMPORARY RELEASE APPLICANTS. THE SELECTION PROCESS ACKNOWLEDGES THAT SOME INMATES WHO SCORE HIGH, SUCH AS SEX OFFENDERS, YOUTHFUL OFFENDERS, AND MENTALLY UNSTABLE APPLICATIONS, MAY BE UNSUITABLE PARTICIPANTS IN TEMPORARY RELEASE. THE DEVELOPMENTS OF THE POINT SYSTEM IS DETAILED IN AN APPENDIX. (DEP)