BACKGROUND: Numerous studies have reported the correlation between AXIN2 polymorphism and cancer risk, but the results seem not consistent. In order to get an overall, accurate and updated results about AXIN2 polymorphism and cancer risk, we conducted this study. METHODS: An updated analysis was performed to analyze the correlation between AXIN2 polymorphisms and cancer risk. Linkage disequilibrium (LD) analysis was also used to show the associations. RESULTS: Seventy-two case-control studies were involved in the study, including 22,087 cases and 18,846 controls. The overall results showed rs11079571 had significant association with cancer risk (allele contrast model: OR = 0.539, 95%CI = 0.478-0.609, PAdjust = 0.025; homozygote model: OR = 0.22, 95% CI = 0.164-0.295, PAdjust< 0.001; heterozygote model: OR = 0.292, 95% CI = 0.216-0.394, PAdjust< 0.001; dominant model: OR = 0.249, 95% CI = 0.189-0.33, PAdjust< 0.001). The same results were obtained with rs1133683 in homozygote and recessive models (PAdjust< 0.05), and in rs35285779 in heterozygote and dominant models (PAdjust< 0.05). LD analysis revealed significant correlation between rs7210356 and rs9915936 in the populations of CEU, CHB&CHS, ESN and JPT (CEU: r = 0.91; CHB&CHS: r = 0.74; ESN: r = 0.62, JPT: r = 0.57), and a significant correlation between rs9915936 and rs7224837 in the populations of CHB&CHS, ESN and JPT (r>0.5), between rs7224837 and rs7210356 in the populations of CEU, CHB&CHS, JPT (r>0.5), between rs35435678 and rs35285779 in the populations of CEU, CHB&CHS and JPT (r>0.5). CONCLUSIONS: AXIN2 rs11079571, rs1133683 and rs35285779 polymorphisms have significant correlations with overall cancer risk. What's more, two or more polymorphisms such as rs7210356 and rs9915936, rs9915936 and rs7224837, rs7224837 and rs7210356, rs35435678 and rs35285779 have significant correlation with cancer susceptibility in different populations.