Obesity has been promoted by a food environment that encourages excessive caloric intake. An understanding of how the food environment contributes to obesogenic eating behavior in different types of individuals may facilitate healthy weight control efforts. In this study, Ecological Momentary Assessment (EMA) via palmtop computers was used to collect real-time information about participants' environment and eating patterns to predict overeating (i.e., greater than usual intake during routine meals/snacks, and eating outside of a participant's normal routine) that could lead to weight gain. Thirty-nine women (BMI = 21.6 ± 1.8; age = 20.1 ± 2.0 years; 61% white) of normal weight (BMI 18.5-25) completed the Three Factor Eating Questionnaire and the Power of Food Scale (PFS), and carried a palmtop computer for 7-10 days, which prompted them to answer questions about eating events, including a count of the types of good tasting high-calorie foods that were available. None of the self-report measures predicted overeating, but BMI interacted with the number of palatable foods available to predict overeating (P = 0.035). Compared to leaner individuals who reported a relatively low frequency of overeating regardless of the availability of palatable food, the probability of overeating among heavier individuals was very low in the absence of palatable food, but quickly increased in proportion to the number of palatable foods available. Our findings suggest that the eating behavior of those with higher relative weights is susceptible to the presence of palatable foods in the environment. Individuals practicing weight control may benefit from limiting their exposure to good tasting high-calorie food in their immediate environment.