This pilot study was attempted to explore the relationships among cognitive load, visual behavior, and reading comprehension in augmented reality (AR) learning environment. An ASL MobileEye-XG eye-tracking system was used to record the participants' visual behaviors while they read Taiwan traditional folk belief information demonstrated by AR technologies. A survey was used to examine participants' cognitive loads due to different cognitive elements (videos, texts and pictures) designed in the AR contents. Currently, the results showed that participants with higher reading comprehensions seemed to have higher cognitive loads from videos. Although no significant relationships were found between visual behaviors and reading comprehensions, some significant correlations were found between visual behaviors and cognitive loads, and also between cognitive loads and reading comprehensions. More results will be discussed in the conference.