Lay summary: Human and animal studies have shown that chronic stress interfers with both homeostatic and hedonic appetite control. Here, we investigated the effect of chronic stress on food preferences and eating behavior in real life settings. In random order, fifty healthy students participated in two test periods of 4-5 days; a stressful period (one week prior to an examination) and a nonstressful period (four weeks after an examination). Food preferences were assessed by counting money spent on highly rewarding foods bought with gift certificates, and changes in eating behavior was further assessed by the Three Factor Eating Questionnaire. Cohen's Perceived Stress Scale, the Recovery-Stress Questionnaire, heart rate variability and Cortisol awakening response were used to evaluate the level of stress. Data on glycemic control, blood pressure, physical activity and sleep were also collected. Forty-four subjects had complete data on the primary outcome. Self-perceived stress was higher and recovery lower in the exam period (p ≤ .001). Subjects were less cognitively restrained (p = .037), less moderately-to-vigorously and lightly physically active (p ≤ .037) and were more sedentary (p = .009) in the examination period. However, no difference was found in money spent on high reward foods, disinhibition or hunger between the examination and control condition. Furthermore, no differences in the physiological markers of stress, glycemic measures and sleep were found. Data does not convincingly support the hypothesis that perceived stress increases the preference for highly palatable foods or leads to adverse effects on different markers of health. However, the stressor might have been to mild to induce obesogenic behaviors.