In the current study our purpose is to investigate the influence of using mathematical communication and whole-class discussions on the student\'s mathematical reasoning ability. In this paper, we report an experiment that promotes whole-class discussions when posing open-ended problems as a tool for helping students learn mathematics for improving their reasoning ability. A school was randomly selected among high schools and two classes were randomly selected among grade 9th classes of this school. These classes were randomly assigned into one of two conditions: whole-class discussion (the experimental group, 27 students) vs. traditional instruction (the control group, 30 students). Both groups were exposed to the same learning materials, solved the same mathematical problems, and were taught by the same experienced teacher. Analysis of covariance showed a significant interactive effect of method of teaching on student reasoning ability.