This study aimed to investigate whether reducing task difficulty by segmenting a complex geometric proof would have a differential influence on experts’ and novices’ cognitive load and reading comprehension. Based on mathematical structures and the theory of reasoning with organization (Duval, 1998), two versions (i.e., segmented and nonsegmented) of a print-based geometric proof were created. Forty-nine experts (i.e., 28 pre-service and 21 in-service math teachers) and sixty-six novices in their eighth-grade year were randomly assigned to either a segmented or non-segmented group. Results showed that for both experts and novices, segmentation helped increase their reading willingness and to lower their perceived task difficulty and cognitive demandingness of task. Segmenting or not, however, would not make a statistical difference with respect to participants’ reading comprehension. Additionally, irrespective of versions of text read, experts’ reading willingness and confidence level were significantly higher than those of the novices. Experts’ perceived task difficulty and cognitive demandingness of task were significantly lower than those of the novices. Also, experts were found to comprehend the geometric proof significantly better than novices. Based on results of the study, future research and instructional strategies for teaching geometric proofs were proposed.