Ischemic stroke is thought to have a polygenic basis, but identification of stroke susceptibility genes and quantification of associated risks have been hampered by conflicting results from underpowered case-control studies. We performed a meta-analysis of all candidate gene association studies in ischemic stroke. Electronic databases were searched up until January 2003 for all case-control and nested case-control studies in English-language journals relating to the investigation of any candidate gene for ischemic stroke in humans. Cases were required to have neuroimaging evidence of the diagnosis. To maintain genetic homogeneity, only studies in white adults were included. Studies that evaluated quantitative traits or intermediate phenotypes were excluded. Data from 120 case-control studies were included. Pooled odds ratios (ORs) with 95% confidence intervals (CIs) from random- and fixed-effects models were calculated. Of 32 genes studied, 15 polymorphisms were identified for which at least 3 studies had been conducted. Statistically significant associations with ischemic stroke were identified for factor V Leiden Arg506Gln (OR, 1.33; 95% CI, 1.12-1.58), methylenetetrahydrofolate reductase C677T (OR, 1.24; 95% CI, 1.08-1.42), prothrombin G20210A (OR, 1.44; 95% CI, 1.11-1.86), and angiotensin-converting enzyme insertion/deletion (OR, 1.21; 95% CI, 1.08-1.35). These were also the most investigated candidate genes, including 4588, 3387, 3028, and 2990 cases, respectively. No statistically significant association with ischemic stroke was detected for the 3 next most investigated genes (factor XIII, apolipoprotein E, and human platelet antigen type 1). There is a genetic component to common stroke. No single gene with major effect was identified; rather, common variants in several genes, each exerting a modest effect, contribute to the risk of stroke. These findings have important implications for the design of future genetic studies and for predictive genetic testing for stroke and other multifactorial diseases.