ObjectiveWe sought to evaluate the accuracy of quantitative three-dimensional (3D) CT angiography (CTA) for the assessment of coronary luminal stenosis using digital subtraction angiography (DSA) as the standard of reference.MethodTwenty-three patients with 54 lesions were referred for CTA followed by DSA. The CTA scans were performed with 256-slice spiral CT. 3D CTA were reconstructed from two-dimensional CTA imaging sequences in order to extract the following quantitative indices: minimal lumen diameter, percent diameter stenosis (%DS), minimal lumen area, and percent area stenosis (%AS). Correlation and limits of agreement were calculated using Pearson correlation and Bland–Altman analysis, respectively. The diagnostic performance and the diagnostic concordance of 3D CTA-derived anatomic parameters (%DS, %AS) for the detection of severe coronary arterial stenosis (as assessed by DSA) were presented as sensitivity, specificity, diagnostic accuracy, and Kappa statistics. Of which vessels with %DS >50% or with %AS >75% were identified as severe coronary arterial lesions.ResultThe correlations of the anatomic parameters between 3D CTA and DSA were significant (r = 0.51–0.74, P < 0.001). Bland–Altman analysis confirmed that the mean differences were small (from −1.11 to 27.39%), whereas the limits of agreement were relatively wide (from ±28.07 to ±138.64%). Otherwise, the diagnostic accuracy (74.1% with 58.3% sensitivity and 86.7% specificity for DS%; 74.1% with 45.8% sensitivity and 96.7% specificity for %AS) and the diagnostic concordance (k = 0.46 for DS%; 0.45 for %AS) of 3D CTA-derived anatomic parameters for the detection of severe stenosis were moderate.Conclusion3D advanced imaging reconstruction technique is a helpful tool to promote the use of CTA as an alternative to assess luminal stenosis in clinical practice.