Excessive exercise has been found to have both negative physical and psychological health effects, with stress being one of the major pathways linking exercise to health. However, so far it is unclear whether those effects are direct effects of exercise behavior or rather mediated by exercise attitudes. Hence, the current study set out to characterize the association between exercise attitudes and stress measures, considering eating attitudes as well as gender as important moderators. Methods: 54 healthy students (33 female) were asked to complete questionnaires assessing eating and exercise attitudes, eating and exercise behaviors, as well as perceived chronic stress levels. Two days of at home saliva sampling were used to assess basal cortisol profiles. Results: No significant associations were found between self-report data and cortisol. A significant main effect of eating attitudes on perceived stress was found (independent of food intake), beta = 2.59, p = .002, while no main effect of exercise attitudes on perceived stress was found, beta = 0.01, p = .89. Further, significant interaction effects of both eating attitudes (independent of food intake), and exercise attitudes (independent of exercise behavior) on perceived stress were found, beta = -4.17, p = .045, beta = -0.42, p = .025 respectively. Conclusion: This exploratory study suggests eating attitudes do affect perceived stress outside of eating behavior. Furthermore, in support of trends found in clinical populations (eating disorders, muscle dysmorphia), it was found that having more positive eating attitudes acts as a protective factor against perceived stress for females, and that having a higher commitment to exercise acts as a protective factor against perceived stress for males.