The present study assessed 64-slice computed tomographic accuracy to quantify minimal lumen area (MLA) and determine lesion severity in intermediate stenosis by angiography compared with intravascular ultrasound (IVUS). Sixty-four-slice computed tomography (CT) has been shown to be effective in coronary stenotic assessment by visual estimation compared with angiography. However, angiography is not an accurate gold standard for intermediate stenotic quantification compared with IVUS. Forty patients (54 lesions) with 30% to 70% coronary stenosis by angiography in a major coronary branch were included. All patients underwent quantitative angiography, retrospective electrocardiographically gated 64-slice CT (Siemens), and IVUS (40-MHz Atlantis; Boston Scientific). MLA was manually traced by 2 blinded and independent operators on 64-slice computed tomographic cross-sectional reconstruction and compared with IVUS MLA. A lesion was considered significant if the MLA was <or=6 mm(2) for the left main coronary artery and <or=4 mm(2) for another epicardial vessel with CT and IVUS. The correlation between IVUS MLA and computed tomographic MLA was r = 0.88 (p <0.001). Interobserver variabilities (mean +/- SD) were 1.2 mm(2) for CT and 1.1 mm(2) for CT versus IVUS. Bland-Altman analysis showed a 95% confidence interval of -42% to +44% for computed tomographic measurement using IVUS as a reference. Sensitivity, specificity, accuracy, and Cohen's kappa coefficient for significant lesion classification using CT were 87%, 72%, 80%, and 0.6, respectively (p <0.0001). In conclusion, when using MLA, 64-slice CT was able to quantify coronary stenosis with good correlation compared with IVUS and determine lesion severity in patients with intermediate lesions by angiography.