ABSTRACT: Relationship between Toll-like receptor-2 (TLR2) and cancer risk has been illustrated in some studies, but their conclusions are inconsistent. Therefore, we designed this meta-analysis to explore a more accurate conclusion of whether TLR2 affects cancer risks. Articles were retrieved from various literature databases according to the criteria. We used STATA to calculate the odds ratio (OR) and 95% confidence interval (95% CI) to evaluate the relationship between certain polymorphism of TLR2 and cancer risk. Finally, 47 case-control studies met the criteria, comprising 15851 cases and 21182 controls. In the overall analysis, people are more likely to get cancer because of -196 to -174del in TLR2 in all five genetic models, B vs. A (OR = 1.468, 95% Cl = 1.129-1.91, P=0.005); BB vs. AA (OR = 1.716, 95% Cl = 1.178-2.5, P=0.005); BA vs. AA (OR = 1.408, 95% Cl = 1.092-1.816, P=0.008); BB+BA vs. AA (OR = 1.449, 95% Cl = 1.107-1.897, P=0.007); BB vs. BA+AA (OR = 1.517, 95% Cl = 1.092-2.107, P=0.013). Meanwhile, rs4696480 could significantly increase the risk of cancer in Caucasians, furthermore, rs3804099 significantly decreased cancer risk in overall analysis, but more subjects are necessary to confirm the results. All in all, this meta-analysis revealed that not only -196 to -174del increased the risk of among overall cancers, Caucasians are more likely to get cancer because of rs4696480, while rs3804099 polymorphism could reduce the risk of cancer in some genetic models. There is no direct evidence showing that rs5743708, rs3804100 and rs1898830 are related to cancer.