OBJECTIVE. The purpose of this study was to investigate the effects of vascular attenuation on the accuracy of stenosis evaluation with 64-MDCT coronary angiography. MATERIALS AND METHODS. A pulsating cardiac phantom was used to simulate the beating heart and coronary arteries of 5 and 3 mm in diameter with three degrees of stenosis (25%, 50%, and 75%) at a heart rate of 55 beats per minute. Coronary vascular enhancement had four attenuation levels: low, 200 H; moderately low, 300 H; moderately high, 350 H; and high, 500 H. Cardiac scans were obtained with 64-MDCT. Percentage stenosis, plaque area, and plaque density were measured on axial images. RESULTS. For 50% and 75% stenosis in 5-mm vessels, there were no significant differences among the four attenuation groups. For 50% and 75% stenosis in 3-mm vessels, significant underestimation of percentage stenosis occurred in the high-attenuation group compared with the moderate-and low-attenuation groups ( p < 0.05). For 25% stenosis in 5-mm vessels, low attenuation led to significant overestimation of degree of stenosis compared with the moderate and high attenuation levels ( p < 0.05). None of the instances of 25% stenosis in 3-mm vessels were detected in the high-attenuation group. Underestimation was found only for 3-mm vessels. For 75% stenosis, all plaques were detected irrespective of contrast attenuation and vessel size. CONCLUSION. Use of higher attenuation leads to a significant underestimation of stenosis in smaller vessels. Lower attenuation leads to slight and clinically acceptable overestimation of stenosis. The optimal vascular attenuation for stenosis detection in coronary 64-MDCT angiography is approximately 350 H.