Results revealed that the participants had an average LSI score of 15.5, compared to average scores ranging from 20.9 to 25.1 for samples of incarcerated adult males from the same jurisdiction. Data on the three outcome measures were collected over a 39-month period revealed that the use of LSI scores for prediction produced relative improvement over chance scores for 1-year and 2-year recidivism, parole failure, and halfway house noncompliance ranging from 54.5 percent to 71.5 percent. Across all types of discharge, offenders defined by the LSI as high risk were consistently more likely to fail on release than were low-risk offenders. Findings indicated that risk assessment by the LSI can be incorporated in a systematic way into criminal justice decisionmaking for female offenders. The identification of low-risk female offenders would permit the use of more humane and cost-efficient alternatives to institutionalization and enable the direction of resources toward intensive rehabilitation for high-risk offenders. Tables, note, and 24 references