forty-eight overweight females were randomly assigned into four groups: 1) stimulus control, 2) exercise, 3) stimulus control plus exercise, and 4) counting calories (control), to determine which treatment was most effective for weight and fat loss. locus of control was assessed with rotter's i-e and saltzer's weight locus of control scales. weekly measures of food intake and energy expenditure were taken. the relation of program adherence to weight loss was also assessed. analyses of covariance showed that there was a significant effect of stimulus control procedures on weight and fat, whereas there was no effect of exercise. the weight loss was maintained six months after the end of the program in the groups that had received stimulus control treatment. analyses of variance indicated that food intake decreased in the stimulus control group, whereas there was no significant effect of exercise on energy expenditure. locus of control orientation did not correlate with amount of weight loss. program adherence did not correlate with amount of weight loss.