THE DEVELOPMENT OF THE PREDICTION (BASE EXPECTANCY) TABLES WAS BASED ON DATA ON 650 INMATES DISCHARGED FROM 3 PRISONS BETWEEN NOVEMBER 1969 AND MAY 1970. OF THESE INMATES, 300 WERE PAROLED AND 350 WERE RELEASED UPON EXPIRATION OF THEIR TERMS. A 5-YEAR FOLLOWUP STUDY FOUND THAT 65.3 PERCENT OF THE PAROLEES AND 80.6 PERCENT OF RELEASEES HAD BEEN REARRESTED, AND THAT 48.5 PERCENT OF THE PAROLEES AND 6.53 PERCENT OF THE RELEASEES HAD BEEN REINCARCERATED. SUCCESS RATES WERE CALCULATED FOR VARIOUS CRITERIA; I.E., VARIOUS LENGTHS OF TIME BETWEEN RELEASE AND REARREST OR REINCARCERATION. TWENTY VARIABLES (E.G., MEMBERSHIP IN GANGSTER ORGANIZATIONS, PAST INCARCERATIONS, EDUCATIONAL BACKGROUND, WORK RECORD, MARITAL STATUS) WERE ALLOTTED POINTS ACCORDING TO EACH OF THE CRITERIA, AND TABLES WERE CONSTRUCTED. THE TABLES USING REINCARCERATION (AS OPPOSED TO REARREST) AS THE CRITERION WERE FOUND TO HAVE GREATER PREDICTIVE POWER. MAJOR FINDINGS REGARDING RECIDIVISM PATTERNS AMONG THE TWO AGE GROUPS OF INMATES ARE SUMMARIZED. DIRECTIONS FOR FURTHER RESEARCH IN THE DEVELOPMENT OF PREDICTION TABLES ARE SUGGESTED. NO SUPPORTING DATA ARE INCLUDED. (LKM)