The analysis used data from 12,526 offenders, of whom 338 were absconders. Results showed that the risk instrument was fairly accurate in identifying offenders likely to become absconders. Probation absconders were more likely to be scored at their initial assessment as not being motivated to change, having frequent address changes, unemployed, younger, and having histories of prior probations and probation revocations. For parolees, the number of prior periods of parole, the number of revocations, and the number of convictions were the only factors showing significant differences between absconders and nonabsconders. Findings indicated the usefulness of risk assessment and the desirability of modifying the risk instrument to reflect the study's results. Research recommendations, tables, and copy of risk instrument.