This study aimed to analyze the role of computer-assisted instruction (CAI) on students' achievement concerning the subject of 'probability'. The experimental pre-and post-test with control group research design was carried out with 48 seventh grade students by conducting The Probability Achievement Test (PAT) to all groups. Data were analyzed by employing an analysis of covariance (ANCOVA) on post-test scores with a pre-test as the covariate and by calculating effect size values. The results revealed that the CAI was more effective in helping the students develop the probability concepts than traditional instruction (TI). Specifically, this study highlights that the CAI tasks that designed for teaching probability were powerful and useful for students to enhance their understanding of important concepts of mathematics and might be used as a new and better way of teaching probability.