Monitoring mental fatigue has become important for improving cognitive performance and health outcomes especially for older adults. Previous models using eye-tracking data allow inference of fatigue during cognitive tasks, such as driving, but they require us to engage in specific cognitive tasks. A model capable of inferring fatigue in natural-viewing situations when individuals are not performing cognitive tasks would help monitor mental fatigue in everyday situations. Moreover, although eyetracking measures exhibit age-related changes, previous models were mainly tested by user groups that did not include older adults. Here, we present a fatigue-detection model including (i) novel feature sets to better capture mental fatigue in naturalviewing situations and (ii) multiple fatigue-detection classifiers of each estimated age group to make it robust to the target’s age. To test our model, we collected eye-tracking data from younger and older adults as they watched video clips before and after performing cognitive tasks. Our model improved accuracy by up to 22.3% compared with a model based on the previous studies, and it achieved 99.4% accuracy. Furthermore, after it was trained using the eye-tracking data before and after cognitive tasks, our model could detect increased mental fatigue of full-time workers after their work with 92.6% accuracy.