The authors examined the relationship between calcified coronary atherosclerosis and an array of cardiovascular risk factors in sequential logistic models to determine the extent to which these markers overlap in their identification of patients at risk for developing coronary heart disease. The prevalence of coronary artery calcium using electron beam computed tomography was 19.4% in this cross-sectional study of a prospective, consecutive, screening cohort of 1999 healthy United States Army personnel (aged 39-50 years). The proportion of the total variance of coronary artery calcium explained by sequential logistic models incorporating conventional, emerging, hereditary, lifestyle, and psychosocial cardiovascular risk variables increased progressively from 9.7% to 14.5%. The best-fit logistic model for the prediction of coronary artery calcium identified age, male gender, Framingham risk score, total cholesterol, high-density lipoprotein cholesterol, triglycerides, smoking, a family history of coronary heart disease, white race, physical inactivity, and lower depression scores as significant independent correlates of coronary artery calcium. These data indicate that the explanatory power of models for atherosclerosis can be significantly improved with the use of emerging, heredity, lifestyle, and psychosocial factors. The large residual variance, however, supports the potential of atherosclerosis imaging to incrementally and independently identify coronary heart disease risk.