Schachter's externality hypothesis suggests that overweight individuals are more likely to be induced to eat by salient external cues than normal weight individuals. While a range of studies have demonstrated the plausibility of this hypothesis in the case of sensory stimuli (e.g., taste cues), there is little evidence that the hypothesis applies to social stimuli. The current study examines this latter proposition by exposing male and female, overweight and normal weight subjects to a same-sex or opposite-sex peer model. Under the guise of engaging in a taste experiment, the subjects were either exposed to a model who tasted experiment, the subjects were either exposed to a model who tasted no crackers (no eat), one cracker (low eat), or twenty crackers (high eat). In addition, control model-absent conditions were also run for purposes of establishing baseline eating rates. If the externality hypotheses were to prevail in social domains, one would expect overweight subjects to be more prone to model the cracker-eating behavior of the peer than normal weight individuals. However, the findings indicate that all subject groups regardless of weight evidence a rather clear modeling effect and all subjects evidence social inhibition effects on their eating behavior as well. Several intriguing interactions among subject sex, model sex, subject weight, and social condition were also found. The discussion explores the relevance of an externality model of overweight eating in social domains, and focuses upon the interesting and somewhat distinct pattern of socially mediated eating exhibited by overweight females.