ABSTRACT: Genetic variants in cytochrome P450 (CYP), platelet membrane receptor (P2Y12, P2Y1), and glycoprotein IIIa (GPIIIa) genes are associated with the efficacy of clopidogrel and adverse clinical events on ischemic stroke (IS) patients. However, few studies have assessed whether gene-gene interactions among these genes influence the risk of IS. The aim of the present study was to investigate the association of fifteen variants with IS and to determine whether these gene-gene interactions increase the risk of IS.Fifteen variants in CYP3A4, CYP3A5, CYP2C8, CYP2C9, CYP2C19, P2Y12, P2Y1 and GPIIIa genes were examined using mass spectrometry methods in 396 patients with IS and 378 controls. Gene-gene interactions were analyzed using generalized multifactor dimensionality reduction (GMDR) methods.Single-gene variant analysis showed no significant differences in the genotype distributions of the fifteen variants between IS patients and controls using the single-locus analytical approach. However, GMDR analysis showed a significant gene-gene interaction among rs17110453A>C, rs2317676A>G, and rs16863323C>T, which scored 10 for cross-validation consistency and 9 for the sign test (P = 0.016). Logistic regression analysis showed that high-risk interactions among rs17110453A>C, rs2317676A>G, and rs16863323C>T were independent risk factor for IS after adjusting for age, hypertension, diabetes mellitus, and hemoglobin A1C (OR=2.24, 95% CI: 1.17-5.62, P=0.005).The rs17110453A>C, rs2317676A>G, and rs16863323C>T three-loci interaction may confer a higher risk for IS. The combinatorial analysis used in this study may be helpful to elucidate complex genetic risk factors for IS.