Adolescents constitute a large proportion of youth in foster care, but few foster care models specifically concentrate on their developmental or treatment needs. This study examined outcomes for an agency-developed, theoretically based treatment foster care (TFC) model using de-identified administrative data from 612 youth discharged from care over a 3-year period using a structural equation modeling path analysis. Results indicate that youth enrolled in Pressley Ridge Treatment Foster Care (PR-TFC) improve in functioning from entry to discharge and that, overall, youth in PR-TFC treatment typically improve in day-to-day functioning as they age, and youth who enter the PR-TFC homes at an older age generally function better at discharge than younger youth who stayed comparable amounts of time. Additional results indicate that older youth, youth with a greater number of clinical diagnoses, youth with a lower pre-test Child and Adolescent Functional Assessment Scale (CAFAS) score, and youth of color spent more time in care, while older youth, youth with fewer clinical diagnoses, youth with lower pre-test CAFAS scores, and European American youth had lower post-test CAFAS scores, indicating better functioning at discharge. This article also provides an example of how administrative data can be used to develop practice-based evidence for a treatment model.