To investigate the effectiveness of role-play methods in education. Meta-analysis was conducted, and publication bias test and heterogeneity test were performed. Robustness was also tested by the removal-by-removal method. Finally, subgroup analyses were conducted to explore the effects on teaching effectiveness under different moderating variables. Twelve articles with 22 effect size samples involving 907 participants, including 821 university students and 86 secondary school students, were finally included. Meta-analysis demonstrated that teaching with the role-play method had a more significant positive effect on students than the control group (ES=0.818, 95% confidence interval 0.600-1.035). We also found that the role-play method had the most significant impact on students' Skills. This study demonstrates that role-play can enhance learning in all dimensions by allowing students to take on different roles and situations.