This evaluation of Florida's extensive system (from 140 to 395 youths) of self-contained juvenile training schools (TS) profiles the delinquent client population and examines the recidivism rate. After discussing the dangers of using recidivism as the sole indicator of rehabilitation, the study's methodology is explained. A random sample follows the subjects from furlough (parole) through a year of followup study. Case files provide data on the child and family; only the background variables of sex, number of prior commitments, and type of offense show a significant relation to success, while in-program experience shows non-significant relation to success. A comparison is made with earlier data, and goals for future studies are suggested. (LS)