In this study, we explored the structural relationship between the students' perceived use of cognitive and metacognitive reading strategies (CMRS) and their reading comprehension of geometry proof (RCGP), and we also examined the differences in students' perceived use of reading strategies among the poor, moderate and good comprehenders. A sample of ninth graders (N = 533) completed a RCGP test and then the CMRS questionnaire. In the exploratory factor analysis with one subsample (n = 150), principal component analysis was used to extract factors of CMRS use for improving the CMRS instrument. Another subsample of students (n = 370) participated in the study on the confirmatory factor analysis with structural equation modelling method. Results revealed that the use of metacognitive reading strategies exerts an executive function over that of cognitive reading strategies, which directly influenced students' RCGP. Our interesting findings were that good comprehenders tended to employ more metacognitive reading strategies for planning and monitoring comprehension and more cognitive reading strategies for elaborating proof compared with the moderate comprehenders, who in turn employed these strategies more often compared with the poor comprehenders.