Eye-tracking measurements may be used as a method of identifying users' actual behavior in a hypermedia setting. In this research, an eye-tracking experiment was conducted in order to validate the construct of cognitive style as a personalization parameter in adaptive e-Learning systems. The main research question was whether the verbalizer/ imager axis of the Cognitive Style Analysis theory reflects actual preferences in an e-learning environment and properly identifies learner types. The findings from a sample of 21 participants reveal statistically significant differences among types of learners; as hypothesized, imagers concentrate on visual content, verbalizers on text, while intermediates are placed in between.