Abstract Introduction Recently, researchers in a number of studies have explored the association between the Toll-like receptor 2 (TLR2) Arg753Gln polymorphism and sepsis risk. However, the results were conflicting. In this meta-analysis, we aimed to confirm the effect of the TLR2 Arg753Gln polymorphism on sepsis risk. Methods Relevant records up to 1 June 2015 were retrieved from the PubMed, Embase, and Web of Knowledge databases. The odds ratios with their corresponding 95 % confidence intervals were used to assess the association between the TLR2 Arg753Gln polymorphism and sepsis risk. The selection of a fixed or random effects model was made according to a heterogeneity test in total and subgroup analyses. Sensitivity analysis and publication bias test were performed to ensure the reliability of our results. Results A total of 12 studies with aggregate totals of 898 cases and 1517 controls met our inclusion criteria for meta-analysis. There were significant associations between the TLR2 Arg753Gln polymorphism and sepsis risk in overall analyses under two genetic models (the allele comparison and the dominant model). In addition, subgroup analyses based on age group, ethnicity, sepsis type, and source of control also showed a significant effect of the TLR2 Arg753Gln polymorphism on sepsis risk. Conclusions Our present meta-analysis supports a direct effect of the TLR2 Arg753Gln polymorphism on sepsis risk, especially in Europeans. The TLR2 Arg753Gln polymorphism might be used as a relevant risk estimate for the development of sepsis. Studies with larger sample sizes and homogeneous groups of patients with sepsis are required for further analysis.