Warfarin has been shown to accelerate vascular calcification in experimental animals, and possibly humans, through inhibition of the vitamin K-dependent protein matrix gla protein, a potent inhibitor of tissue calcification. We performed a cross-sectional analysis of the extent of coronary artery calcification (CAC) in patients without coronary heart disease, currently taking or referred for warfarin therapy. The primary end point was severity of CAC measured by electron beam computed tomography attributed to duration of warfarin use, after adjustment for cardiovascular risk factors. Seventy patients (46 men, mean age 68 +/- 13 years) were enrolled from three groups of warfarin use duration: (1) <6 months (n = 31, mean duration 1 +/- 1 months), (2) 6-24 months (n = 11), and (3) >24 months (n = 28, mean 67 +/- 40 months). Overall, the mean total CAC score (Agatston) was 293 +/- 560: group 1 (175 +/- 285), group 2 (289 +/- 382), and group 3 (426 +/- 789). In univariate analysis, there was a nonsignificant trend to increased CAC with increasing warfarin exposure (P = 0.18). Bivariate analysis revealed no correlation between warfarin duration and CAC score (r = 0.075, P = 0.537). Linear regression for the independent variable coronary calcium score controlling for warfarin treatment duration and intensity (duration of warfarin use months x mean INR), Framingham risk score, and creatinine clearance showed that only the Framingham risk score was associated with CAC (P = 0.001). Among patients without known coronary heart disease, duration of warfarin exposure was not associated with extent of coronary calcification.