The relationship between stress, restraint, and eating has been studied using various methods, including retrospective self-reports of stress and eating that are open to inaccuracies. Additionally, laboratory research has not systematically varied the fat content and sweetness of food items to assess how stress relates to specific food preferences. In this study of 40 women we examined the role of restraint and experimentally induced stress on the amount of sweet, salty, high-fat, and low-fat food consumed. High-restraint women ate more high-fat food than did low-restraint women, regardless of stress level. High-stressed women preferred sweet, high-fat food more than did low-stressed women, whereas low-stressed women ate more low-fat than high-fat food. There was no interaction between restraint and stress level. Social influence effects of small-group testing may have increased the ego-threat of the stressor or disinhibited high-restraint women in both stress groups. Future laboratory research is needed to assess the role of the presence of others in both stress induction and eating behaviors.