To evaluate the clinical efficacy of moxibustion and drug therapies in treating primary dysmenorrhea(PD).Methods:The randomized controlled trials(RCTs) of moxibustion versus drug therapies for PD were retrieved from the China Biology Medicine(CBM)，China National Knowledge Infrastructure(CNKI)，Chinese Citation Database(CCD)，China Science Periodical Database(CSPD)，PubMed，Embase，and Cochrane Library with the time interval from inception to April 10，2021.The Jadad scale was employed to evaluate the included articles，and the Cochrane's risk of bias assessment tool to evaluate the risk of bias.The meta-analysis was performed in Rev Man5.4 and Stata 15.Results:A total of 16 RCTs were included.There was a high risk in the implementation of the blind method，and the randomization and concealment schemes were unknown.The results of meta-analysis showed that the total response rate differed between the moxibustion group and the drug group(RR=1.25，95%CI of 1.19 to 1.31，P<0.000 01).The left and right side of the funnel plot were not symmetrical，suggesting that there may be publication bias.The cure rate differed between the moxibustion group and the drug group(RR=2.45，95%CI of 2.00 to 3.01，Z=8.56，P<0.000 01).The changes in the dysmenorrhea symptom score and visual analogue scale score also differed between the two groups(P=0.000 01).Conclusion:Moxibustion had better efficacy than drugs in the treatment of PD.However，due to the small number and low quality of the included RCTs，large-sample，multi-center，and high-quality RCTs are needed to evaluate the safety and efficacy of moxibustion.