Circulating osteoprotegerin (OPG) has been shown to be elevated in patients with vascular disease. The role of OPG as a biomarker for atherosclerosis in a large, unselected population is not well known. Plasma OPG levels were measured in 3,386 subjects in the Dallas Heart Study, a multiethnic, population-based probability sample of adults aged 30 to 65 years. Coronary artery calcium (CAC) was measured by electron beam computed tomography. Aortic plaque was assessed by magnetic resonance imaging. Multivariable logistic regression was used to assess associations among OPG, cardiovascular risk factors, CAC, and aortic plaque. Age, female gender, black race, smoking, personal and family history of coronary artery disease (CAD), diabetes mellitus, hyperlipidemia, CAC, and aortic plaque were significantly associated with higher plasma OPG levels (p <0.01) in univariable analyses. The prevalence of CAC and aortic plaque increased across OPG quartiles (p <0.001 for each). An OPG level in the fourth quartile was independently associated with CAC (RR 1.39, 95% confidence interval 1.01 to 1.93) and aortic plaque (RR 1.42, 95% confidence interval 1.09 to 1.86) after adjustment for age, gender, smoking, diabetes, hyperlipidemia, and family history of premature CAD. In conclusion, plasma OPG is independently associated with CAC and aortic plaque in an unselected population, suggesting it may be a novel biomarker for atherosclerosis in humans.