Screening for subclinical atherosclerosis has been advocated for individuals at intermediate global risk for coronary heart disease (CHD). However, the distribution of subclinical atherosclerosis test values across CHD risk strata is unknown. We studied a stratified random sample of 292 participants (mean age 59.5 years, 50% women) from the offspring cohort of the Framingham Heart Study who were free of clinically apparent cardiovascular disease. We assessed abdominal and thoracic aortic plaque burden by cardiovascular magnetic resonance (CMR), coronary artery calcification (CAC) and thoracic aortic calcification (TAC) by electron beam computed tomography, and common carotid intima-media thickness (C-IMT) by ultrasonography. We categorized the upper 20% of each measurement as a high level of atherosclerosis and evaluated these variables across clinically relevant Framingham CHD risk score strata (low, intermediate, and high risk). In age-adjusted analyses in men and women, correlations across CMR aortic plaque, CAC, TAC, and C-IMT were low (maximum r = 0.30 for CAC:TAC in women, p <0.005). In men and women, the proportion of subjects with high atherosclerosis test results for any of these measurements increased significantly across Framingham CHD risk score strata (Kruskal-Wallis test, p <0.0001). In the intermediate Framingham CHD risk score category, 14% of men and 25% of women had a high atherosclerosis result on >or=2 measurements. However, different participants were identified as having high atherosclerosis by each modality. For example, in a comparison of the overlap across CMR aortic plaque, CAC, and C-IMT, only 4% of men and 16% of women were classified as having high atherosclerosis on all 3 measurements. In conclusion, in a community-based sample, correlations among subclinical atherosclerosis test results are low, and a substantial proportion has high levels of subclinical atherosclerosis detected on >or=2 imaging tests.