Obesity is a major public health problem in America. The majority of U.S. adults are now overweight, and over one-fifth of U.S. adults are obese (Kuczmarski, Carroll, Flegal, & Troiano, 1997). Obesity is associated with increased risk of numerous medical disorders and all-cause mortality (National Heart, Lung, & Blood Institute, 1998) as well as social stigmatization (Wadden & Stunkard, 1996). Behavior therapy has played a large role in the development of effective treatments for obesity. This paper reviews the history of behavioral obesity treatment, the effectiveness of the behavioral approach, and limitations of behavioral weight loss treatments. HISTORY OF THE BEHAVIORAL TREATMENT OF OBESITY Historically, there have been a number of therapeutic approaches to weight loss based on purely medical or physiological formulations. However, a psychodynamic interpretation of obesity became common in the 1950s. According to this model, obesity is related to underlying difficulties with identifying, containing, and expressing thoughts, feelings, and needs (Moreno, Fuhriman, & Hileman, 1995). Treatments based on this formulation focused on helping patients identify and tolerate troublesome experiences that were acted out in eating behavior, not on changing eating and physical activity directly. As might be expected, this approach was not in the treatment of obesity. For example, Stunkard (1958) revealed that only 12 of 100 patients treated in a specialty clinic of a hospital lost more than 20 pounds. Furthermore, only two of the 12 successful patients maintained their weight loss after two years. Stunkard summed up the state of obesity treatment in the 1950s in decidedly pessimistic terms: Most obese persons will not stay in treatment for obesity. Of those who stay in treatment most will not lose weight and of those who do lose weight, most will regain it. As part of the reaction to psychoanalysis and psychoanalytic theory, in the early 1960s behavioral theorists began to conduct weight loss treatments based on learning principles. These early behavioral weight loss studies were more concerned with testing the efficacy of behavioral theory than treating obesity per se; weight loss was merely a convenient outcome variable (Brownell & Jeffery, 1987; Brownell & Kramer, 1989). Ferster, Nurenberger, & Levitt (1962/1996) were the first to apply learning theory to the treatment of obesity. Ferster and colleagues conceptualized obesity as a failure of self-control in which overweight individuals focus on the immediate reinforcing consequences of eating (e.g., taste of food) as opposed to the ultimate aversive consequences of overeating (e.g., increased weight and body fat). Based on this formulation, a major aspect of Ferster et al.'s treatment involved amplifying the aversive consequences of overeating. Patients developed an extensive verbal repertoire of the negative consequences of eating to be used in situations associated with eating. By pairing eating with the ultimate aversive consequences of overeating, it was thought that undesirable foods and situations would become conditioned aversive stimuli. Ferster et al.'s treatment also involved stimulus control in which patients identified places, times, and events which they usually ate as well as people with whom they commonly ate. Patients were taught to break these connections by eating regular meals at one location (e.g., the dinner table) and to do nothing else while eating (e.g., reading or watching television). Additionally, chaining was used, in which patients lengthened the number of behaviors required to eat. Specific chaining techniques included waiting until the first bite was finished before taking another bite, putting the fork down between bites, and keeping food in hard to reach places. Ferster and colleagues (1962/1996) did not provide outcome data on their program; however, this first application of learning principles to obesity appears to have been largely unsuccessful. …