Abstract With recent research supporting a cognitive explanation of overeating, two experiments were designed to test the theory that the anticipated deprivation associated with dieting may lead to overeating once restraint is broken. In each of 2 experiments, 60 female undergraduates were preloaded and randomly assigned to no deprivation and deprivation conditions. The latter group anticipated a 4-hour food deprivation between two experimental sessions, while the former did not. The amount of food consumed during a 15-minute free eating period served as the dependent measure. Restraint scale scores were used to divide subjects into low-restraint and high-restraint groups. In Experiment 1, deprivation subjects ate significantly more than no deprivation subjects, F(1, 56) = 5.02, p < .05; in Experiment 2, high-restraint subjects ate significantly more than low-restraint subjects, F(1, 56) = 5.07, p < .05. The possible role of contextual cues is explored, and implications for weight-reduction programs are outlined.