A number of studies have investigated the association between tumor necrosis factor (TNF)-α gene polymorphisms and ischemic stroke susceptibility. However, results of different individual studies are often inconsistent. To provide a more robust evaluation of the association between polymorphisms of the TNF-α gene and ischemic stroke risk, we performed a systematic review with multivariate meta-analyses. PubMed, Embase, CNKI, and WanFang databases were searched up to December 20, 2014. Two reviewers independently extracted information and assessed quality of included studies after all the eligible studies were identified. Afterward, multivariate meta-analyses were performed using Stata 13. The estimation of polymorphisms and disease risk was presented by odds ratios (ORs) and corresponding 95 % confidence intervals (CIs). Forty-nine eligible case-control studies from 25 articles that explored the association between 10 TNF-α polymorphisms and ischemic stroke were indentified from aforementioned databases. The results of multivariate meta-analysis showed a significant association between -238G/A polymorphism (4760 patients and 4389 controls) and ischemic stroke risk in heterozygotes compared with wild genotype (AG vs. GG: OR 1.44, 95 % CI 1.11-1.87; AA vs. GG: OR 1.98, 95 % CI 0.73-5.40). No significant association of -308G/A, -857C/T, and -1031T/C polymorphisms was observed. The results of stratification analyses of -238G/A polymorphism showed that the AG genotype only increased the risk of ischemic stroke in Asians compared to GG genotype. No additional significant association was observed in this study. In conclusion, the present systematic review and meta-analysis support a prominent role of the TNF-α -238G/A polymorphism in the risk of ischemic stroke in Asian adults only, but do not support the role of -308G/A, -857C/T, -1031T/C, -244G/A, -367G/A, -646G/A, -806C/T, -863C/A, and +448G/A in the risk of ischemic stroke. The current evidence warrants further studies with high quality and large sample size to confirm.