The use of cardiovascular risk scores remains the foundation for risk stratification to guide clinical management. Clinicians have access to several cardiovascular risk scores in practice settings. While having several risk scores with different risk factors may provide more information, it does not imply accuracy of the cardiovascular risk score used to calculate individual patient cardiovascular risk. The objective of this study was to compare the Framingham Risk score, Reynolds Risk scores, and the Pooled Cohort Risk Equation (3 commonly used equations) scores with respect to ability to predict cardiovascular events in a diverse ethnic population. Additionally, the potential predictive utility of three novel risk factors (carotid intima media thickness, peripheral arterial tonometry and vasa vasorum) was examined in relation to ability to improve 10-year cardiovascular risk prediction. A secondary analysis of the longitudinal prospective study cohort known as Heart Strategies Concentrating On Risk Evaluation (Heart SCORE) was conducted. The cardiovascular risk scores of study participants who did and did not experience a cardiovascular event composite index consisting of myocardial infarction, death, stroke, acute ischemic stroke, or revascularization were assessed using methods of calibration and discrimination overall and by race and gender. When examining performance of the 3 risk scores, the overall 10-year absolute predicted cardiovascular risk varied substantially (e.g. approximately 2-fold) and this wide variation in predicted 10-year cardiovascular risk was present across race and gender. Nonetheless, despite the wide variation in estimates of absolute risk, the 3 cardiovascular risk score equations were strongly associated with future cardiovascular risk overall and by race and gender. There was some indication that the Reynolds risk score was the most accurate measure of future cardiovascular risk. The 3 novel risk factors examined did not significantly improve 10-year cardiovascular risk prediction above and beyond the standard demographic and clinical variables used in these well-known equations.