BECAUSE THE LIKELIHOOD OF FURTHER CRIMINAL CONDUCT IS IMPORTANT IN ALLOCATION DECISIONS FOR PAROLE SUPERVISION RESOURCES, A 50 PERCENT SAMPLE WAS SELECTED FROM THE POPULATION OF ALL FEDERAL PRISONERS WITH MAXIMUM SENTENCES OF MORE THAN A YEAR WHO WERE RELEASED TO THE COMMUNITY IN THE FIRST 6 MONTHS OF 1970. AN INFORMATION SHEET WAS DRAWN CONTAINING OVER 60 ITEMS RELATING TO THE OFFENSE, SENTENCE STRUCTURE, CRIMINAL RECORD, AND OTHER BACKGROUND CHARACTERISTICS. A 6-YEAR FOLLOWUP PERIOD WAS MADE POSSIBLE BY USING RECORDS PROVIDED BY THE FBI. RESEARCH RESULTS INDICATED THAT, AS A GROUP, FEDERAL PRISONERS RELEASED IN 1970 SHOWED A HIGH LIKELIHOOD OF REARREST WITHIN A 6-YEAR FOLLOWUP PERIOD, AND THAT THE LIKELIHOOD OF A NEW ARREST DURING A GIVEN YEAR WAS RELATED TO THE SALIENT FACTOR SCORE AND THE YEAR AFTER RELEASE. A METHOD FOR THE PRACTICAL APPLICATION OF THIS KNOWLEDGE IN DECISIONMAKING FOR PAROLE SUPERVISION IS GIVEN AND PROVIDES EMPIRICAL GUIDANCE FOR DECISIONS ABOUT THE INTENSITY OR DURATION OF SUPERVISION. TABULAR DATA ACCOMPANY THE REPORT AND NOTES AND REFERENCES ARE GIVEN. (MJW)