THE NINE VARIABLES EXAMINED WERE: PREINSTITUTIONAL VARIABLES -- NUMBER OF PRIOR FELONY CONVICTIONS AND TYPE OF LAST FELONY CONVICTION (PROPERTY OR NONPROPERTY); INTRAINSTITUTIONAL VARIABLES -- PARTICIPATION IN VOCATIONAL TRAINING DURING INCARCERATION, PARTICIPATION IN ACADEMIC TRAINING DURING INCARCERATION, OR PROCESSED BY A DIAGNOSTIC CENTER DURING INCARCERATION; AND POSTINSTITUTIONAL VARIABLES -- AGE AT TIME OF PAROLE, HOURLY WAGE RECEIVED ON JOB, NUMBER OF CONTACTS WITH PAROLE OFFICER DURING FIRST 3 MONTHS, AND PAROLE OFFICER TO WHICH ASSIGNED (RURAL OR URBAN). THE NUMBER OF PRIOR FELONY CONVICTIONS ACCOUNTED FOR ALMOST HALF THE TOTAL VARIANCE. ALTHOUGH THERE WAS CORRELATION BETWEEN AGE AT TIME OF PAROLE AND NUMBER OF PRIOR FELONY CONVICTIONS, AGE AT TIME OF PAROLE WAS NOT SIGNIFICANTLY CORRELATED WITH PAROLE OUTCOME, INDICATING THAT FOR THIS SAMPLE PRIOR FELONY CONVICTIONS, NOT AGE, IS THE SIGNIFICANT STATIC VARIABLE. HOURLY WAGE ON PAROLE JOB AND NUMBER OF CONTACTS WITH PAROLE OFFICER DURING THE FIRST 3 MONTHS OF PAROLE WERE ALSO SIGNIFICANT PREDICTORS. IT IS SUGGESTED THAT A PAROLEE WHO IS EARNING AN ADEQUATE INCOME FEELS NO NEED TO RESORT TO CRIME. NONE OF THE THREE REHABILITATIVE PROGRAMS -- ACADEMIC TRAINING, VOCATIONAL TRAINING, AND DIAGNOSTIC EVALUATION -ACCOUNTED FOR MORE THAN A SMALL PERCENTAGE OF THE VARIANCE. EXAMINATION OF THE CASE RECORDS ALSO SHOWED THAT OLDER MEN WITH LENGTHY CRIMINAL HISTORIES TENDED NOT TO HAVE BEEN IN REHABILITATIVE PROGRAMS. IT IS SUGGESTED THAT GREATER EFFORTS BE MADE TO ENROLL REPEAT OFFENDERS IN SUCH PROGRAMS. THIS, COMBINED WITH CLOSE SUPERVISION, MIGHT REDUCE RECIDIVISM. FURTHER SUCH MULTIFACTOR ANALYSES ARE URGED. A BIBLIOGRAPHY IS APPENDED. (GLR)