Because the safety of the community or the safety or health of the defendant are issues in some cases, a presentence evaluation is not always practical within a community setting; therefore, the Federal Bureau of Prisons, the North Carolina Department of Correction, and a number of other correctional systems have instituted presentence evaluation programs within the confines of a prison. Presentence studies within prisons are more costly due to the expense of incarceration and the loss of liberty suffered by the defendants. With this increased cost, prison-based evaluative services must have an effective mechanism for screening referrals, thus ensuring that only those requiring the more costly evaluation are selected for in-prison study. To determine the effectiveness of the North Carolina screening for in-prison evaluations, the records of all 168 defendants admitted to the Presentence Diagnostic Program (in-prison evaluation) in 1974 were used as the study sample. Appropriateness of referral was considered the presence of certain predefined characteristics, suggesting that the well-being of the community or of the defendant would be threatened should the defendant remain in the community during the course of the presentence evaluation. These characteristics were recidivist, prior mental health treatment, unusual or bizarre crime, general behavior unusual, sexual offenders, violent offenders, and mental retardation. The findings suggest that most referrals to the prison-based presentence study are appropriate, with 89 percent of the cases referred presenting one or more of the predefined characteristics. Tabular data are provided.