The hypothesis that negative emotional life states accompanying life stresses are associated with overeating and weight gain cannot be adequately tested solely by laboratory analogue studies. Naturalistic, short-term longitudinal designs are needed in which individuals susceptible to mood eating can be identified prior to a stressful event and followed up to determine weight changes. College students in two samples completed questionnaires on height, weight, restraint, eating habits, self-image, and life events. Mood eating tendencies, which were more frequently reported by both overweight and normal-weight females than males, were associated with self-monitored dysphoric moods and negative self-evaluations. Subjects with high scores on the Mood Eating Scale were likely to be restrictive dieters reporting binge-eating tendencies and dissatisfaction with weight-related appearance. Results suggest that the Mood Eating Scale is a viable measure of a mediating variable, a cognitive appraisal process intervening between life and weight changes. (Author/HLM)