p53 tumour suppressor gene Arg72Pro polymorphism has been associated with endometriosis. However, the current available data were inconsistent. We performed this meta-analysis to estimate the association between p53 Arg72Pro polymorphism and endometriosis. Electronic screening of PubMed library was conducted to select studies. Studies containing available genotype frequencies of Arg72Pro were chosen, and pooled odds ratio (OR) with 95% confidence interval (CI) was used to assess the association. Six published studies, including 749 endometriosis and 857 controls were identified. The overall results suggested that the variant genotypes were not associated with the endometriosis risk (Pro/Pro + Arg/Pro vs. Arg/Arg: OR = 1.552, 95% CI 0.916-2.632, = 0.103). In the stratified analysis, individuals carried the Pro allele in a dominant model had increased risk of endometriosis (OR = 2.595, 95% CI 1.005-6.702, = 0.049) in Asian subjects. The symmetric funnel plot, the Egger's test ( = 0.602), and the Begg's test ( = 0.167) were all suggestive of the lack of publication bias. However, the association was not significant between this polymorphism and endometriosis in Caucasian (OR = 1.005, 95% CI 0.755-1.337, = 0.972). This meta-analysis suggests that p53 codon 72 Pro/Pro + Arg/Pro genotypes are associated with increased risk of endometriosis in Asian. To validate the association between p53 codon 72 polymorphism and endometriosis, further studies with larger participants worldwide are needed.