PURPOSE:Long-term evaluation of an empirically based program for the treatment of morbidly obese, low-income, minority adolescents. METHODS:In the first year, 150 participants received cognitive-behavior therapy, nutritional education, medical monitoring, and structured exercise training. Weights and heights were collected at follow-up. Successful and Less Successful groups were delineated based on change in body mass index z-scores. Medical record review provided weight and height data one year before treatment for a subset of participants (comparison group). Analyses of variance examined differences between groups. Discriminant function analyses examined predictors of success in the domains of socioeconomic status, individual factors and psychological functioning, family factors, and prior weight loss behaviors. Correlates of success were explored. RESULTS:Eighty-three adolescents participated in the follow-up, on average 23 months (SD = 3.8) after initial assessment. Nineteen (23%) participants achieved clinically meaningful weight change (-.70 z-scores or better). Discriminant function analyses, correlations, and direct comparisons showed that the Successful group, compared with the Less Successful group, on average, attended 50% more sessions over 84% more weeks, were heavier initially, had somewhat better critical weight control skills (e.g., self-monitoring) before and during the program, and were somewhat more psychologically and intellectually challenged before treatment. CONCLUSION:Although some of these very high risk adolescents made clinically significant progress, more intensive treatments, like long-term residential treatments and bariatric surgery, may produce even more favorable outcomes.