Conventional wisdom suggests that the Teaching-Family Model (TFM) approach to treating youthful offenders is not effective in reducing post-treatment recidivism. This article reviews two major studies referenced in support of this widespread perception. Data presented in one widely referenced study are treated with a Cochran-Mantel-Haensel test, which, the author argues, is appropriate for data originally presented in two separate 2 X 2 tables (one for boys and one for girls). The construct and statistical conclusion validity of a major evaluation study presented to the NIMH is critically evaluated and discussed. A revised view of the leading TFM evaluations has implications for public policy regarding juvenile justice. The author suggests that a belief in the lack of post-treatment efficacy associated with community-based residential treatment has resulted in harsher treatment of juveniles and a higher incarceration rate. (Contains 2 figures, 3 tables, and 2 footnotes.)