Despite major advances in the diagnosis and treatment of atherosclerotic cardiovascular disease (CVD) in the past century, it remains a serious clinical and public health problem. Multivariable risk factor analysis is now commonly performed to identify high-risk candidates for CVD who need preventive measures and to seek out clues to the pathogenesis of the disease. The set of risk factors used for the former is constrained by practical considerations, and the set of risk factors used for the latter is constrained by the hypothesis being tested. This report reviews the evolution and usefulness of multivariable risk functions crafted for estimating risk of clinical manifestations of atherosclerosis and for gaining insights into their pathogenesis. Decades of evaluation of CVD risk factors by the Framingham Study led to the conclusion that CVD risk evaluation is most fruitfully appraised from the multivariable risk posed by a set of established risk factors. Such assessment is essential because risk factors seldom occur in isolation, and the risk associated with each varies widely depending on the burden of associated risk factors. About half the CVD in the general population arises from the segment with multiple marginal risk factor abnormalities. Although disease-specific profiles are available, a multivariable risk formulation for coronary disease comprised of age, sex, the total/high-density lipoprotein cholesterol ratio, systolic blood pressure, glucose intolerance, cigarette smoking, and electrocardiography-left ventricular hypertrophy is also predictive of peripheral artery disease, heart failure, and stroke because of shared risk factors. Correcting risk factors for any particular CVD has the potential to protect against > or =1 of the others. Multivariable risk stratification is now recognized as essential in efficiently identifying likely candidates for CVD and quantifying the hazard.