The paper describes the effectiveness of a large-scale behavioral program for the treatment of obesity. Results are compared with previous findings in a review of similar interventions. Primary conclusions are: (1) Behavioral interventions produce consistent, but only modest, weight losses; (2) these losses are maintained over a l-year period; and (3) interclient variability is large and cannot be accounted for by known subject or program characteristics. Examination of weight loss curves showed that the rate of weight reduction slowed significantly over time, and weight losses early in treatment were highly related to later weight loss. Further research to identify sources of variability and to elucidate the relationship between behavior change and weight change is encouraged.