This paper describes the outcomes of a study that explored the significance of the Severity of Illness Rating Scale (SIRS) as a tool for utilization review and program evaluation. The SIRS was first introduced in March 1993 and soon became the centerpiece for the Arthur Brisbane Child Treatment Center's Utilization Review Plan. Since its implementation, the monthly average in-house lengths of stay at Arthur Brisbane (an intermediate to long-term state psychiatric hospital) have decreased 142 days, from 234 days in March 1993 to 92 days in February 1996. The SIRS study involved 63 adolescents (ages 11-17) admitted and discharged between August 1, 1994 and July 31, 1995. The average admissions' SIRS scores, discharge settings, and the adolescents' ages proved to be reasonably good predictors of length of stay. Adolescents with higher SIRS scores, who were placed residentially and who were younger, had longer stays. The SIRS scores supported the appropriateness of admissions and, to a lesser extent, program assignments. The SIRS scores also reported improvement from the point of admission to discharge in 7 out of 8 categories of functioning. (CR)