Objective To systematically assess the correlation between smoking and the risk of endometriosis, so as to offer scientific basis to health education and preventing decision. Methods A literature search was performed in The Cochrane Library, Pubmed, Embase, CBM, CNKI and Wanfang database to collect the case control studies on the correlation between smoking and endometriosis. Two reviewers independently screened the literatures according to the inclusion and exclusion criteria, extracted the data, assessed the quality, and then conducted Meta-analyses on the 13 included RCTs by using RevMan 5.0 software, with calculation of the OR value and 95%CI. Results A total of 13 case control studies involving 14260 cases were included, of which 1900 ones were endometriosis. The quality assessment indicated that 2 studies were in quality of Level A, 4 were Level B, 7 were Level C, totally meant low quality. Meta-analyses showed that compared with non-smokers, there was no increasing possibility of endometriosis in smokers (OR= 0.91, 95%CI 0.82 to 1.02). The geographical subgroup analyses showed there was no significant difference in the incidence of endometriosis between the non-smokers and smokers in North America (OR=0.96, 95%CI 0.84 to 1.08), but a significant difference was found between non-smokers and smokers in Europe (OR=0.72, 95%CI 0.54 to 0.97). Conclusion There is no causative relationship between smoking and incidence of endometriosis. However, more high-quality trials are expected for further study because of the heterogeneity and poor quality of the current included studies.