THE STATISTICS SHOW THAT 160 INMATES (17.07 PERCENT OF THE STUDY POPULATION) WERE RETURNED TO CUSTODY. RECIDIVISM RATES WERE HIGHER FOR MALES THAN FOR FEMALES, FOR BLACK MALES THAN FOR WHITE MALES, FOR PERSONS CHARGED WITH PROPERTY CRIMES THAN FOR PERSONS CHARGED WITH CRIMES AGAINST PERSONS, AND FOR PERSONS RELEASED BY MANDATORY CONDITIONAL RELEASE (AS COMPARED TO PAROLEES AND PERSONS WHOSE SENTENCES HAD EXPIRED). THERE WAS VERY LITTLE DIFFERENCE IN RECIDIVISM RATES FOR PERSONS RELEASED FROM VARIOUS TYPES OF FACILITIES (E.G., MAJOR PRISONS, COMMUNITY CENTERS). RECIDIVISTS HAD BEEN RELEASED AN AVERAGE OF 11.7 MONTHS PRIOR TO BEING RETURNED TO CUSTODY. OVER HALF OF ALL RECIDIVISTS WERE RECOMMITTED WITHIN A YEAR. AVERAGE TIME BETWEEN RELEASE AND RECOMMITMENT WAS 10.8 MONTHS FOR PAROLEES, 12.4 MONTHS FOR MANDATORY CONDITIONAL RELEASES, AND 13.9 MONTHS FOR PERSONS WHOSE SENTENCES HAD EXPIRED. OF ALL RECIDIVISTS, 36.9 PERCENT WERE RETURNED FOR TECHNICAL VIOLATIONS OF CONDITIONAL RELEASE AGREEMENTS (PAROLE AND MANDATORY CONDITIONAL RELEASE). AMONG RECIDIVISTS, PERSONS RELEASED AFTER COMMITMENT FOR A PROPERTY OFFENSE TENDED TO BE RECOMMITTED FOR AN OFFENSE AGAINST A PERSON. RECIDIVISM RATES WERE LESS THAN THOSE COMMONLY REPORTED IN THE NEWS MEDIA. OVERALL, SELECTION CRITERIA USED BY THE PAROLE BOARD ARE NO MORE EFFECTIVE THAN THE CRITERIA THAT RESULT IN AN INMATE BEING RELEASED AFTER EXPIRATION OF SENTENCE. THE HIGHEST RISK GROUP APPEARS TO BE INMATES RELEASED FROM MAJOR INSTITUTIONS ON MANDATORY CONDITIONAL RELEASE. DIRECTIONS FOR FURTHER ANALYSIS ARE SUGGESTED. TABULAR DATA AND A BIBLIOGRAPHY ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED--LKM)