OBJECTIVES This study estimated medical expenditures attributable to coronary artery disease (CAD) in the US and investigated CAD case definition in a nationally representative sample. METHODS Data from 2005 from the nationally representative Medical Expenditure Panel Survey (MEPS) were used to estimate the population with CAD and their medical expenditures. CAD cases were identified by patient-reported myocardial infarction (MI), angina pectoris and/or coronary heart disease (CHD). Case definition analysis used multivariable logistic regression. Person-level and national expenditures were estimated on a logarithmic scale using a maximum likelihood Heckman selection model and Smearing re-transformation. All analyses employed Taylor series linearization methods to account for the complex survey design and adjusted for age, race, ethnicity, gender, income, education, and overweight status. RESULTS In the 2005 civilian noninstitutionalized adult population (n= 22262), there were 1016 CAD cases using a strict definition of either MI or angina and 1266 cases using a broad definition including patient-reported CHD. Of those reporting CHD (n=702), only 65% (n=458) had MI or angina . Females and blacks reporting CHD were less likely to have MI or angina (OR=0.52, 0.39, p<0.01). Annual direct medical expenditures attributable to strictly defined CAD were estimated to be $6456 ($2007), on average, per person. Expenditures for broadly defined CAD were $6453 ($2007). Based on this estimate and the weighted estimate of persons with CAD (approximately 9.8 million), the projected annual US medical expenditures attributable to CAD are $63 billion ($2007). CONCLUSIONS While CAD and CHD are generally diagnosed by either MI or angina, patient-reported CHD is not consistent with this definition. Consistency appears to vary with gender and race. While the CAD case definition varies, expenditures do not. Results of this study indicate that direct medical expenditures associated with CAD in the US are substantial.