It is unresolved whether elevated homocysteine in coronary artery disease (CAD) is the cause of arteriosclerosis or its consequence. In contrast, genetic variants of enzymes that metabolize homocysteine cannot be altered by arteriosclerosis. Consequently, their association with CAD would permit to imply causality. We modeled by regression analysis the effect of 11 variants in the methionine cycle upon CAD manifestation in 591 controls and 278 CAD patients. Among the examined variants only the carriership for the c.844ins68 in the cystathionine beta-synthase (CBS) gene was associated with a significantly lowered risk of CAD (OR=0.56; 95% CI=0.35-0.90 in the univariable, and OR=0.41, 95% CI=0.19-0.89 for obese people in the multivariable analysis, respectively). Healthy carriers of the c.844ins68 variant exhibited, compared to the wild type controls, significantly higher postload ratios of blood S-adenosylmethionine to S-adenosylhomocysteine (61.4 vs. 54.9, p=0.001) and of plasma total cysteine to homocysteine (8.6 vs. 7.3, p=0.004). The changes in these metabolites are compatible with an improved methylation status and with enhanced activity of homocysteine transsulfuration. In conclusion, the coincidence of clinical and biochemical effects of a common c.844ins68 CBS variant supports the hypothesis that compounds relating to homocysteine metabolism may play role in the development and/or progression of CAD.