Participating in a laboratory study of the effects of stress on eating, healthy, nonsmoking men and women were shown a film about industrial accidents (stress condition) or a pleasant travelogue (control condition) and had access to sweet, salty, and bland snack food. Analyses of food consumption data revealed a significant interaction between sex of subject and the stress manipulation: Stress markedly and significantly decreased food consumption by men but resulted in some increased food consumption by women. Across the three food taste categories, men consistently ate less under stress than they did in the control condition. In contrast, women ate nearly twice as much sweet food and more bland food under stress than they did in the control condition, but these effects were not statistically significant. These results indicate that the relationship between stress and eating depends on the sex of the subject and may relate to type of food available. Contrasting generalizations about stress and eating reported in the archived literature can largely be reconciled by considering these interacting variables.