The efficacy of contrast-enhanced multislice computed tomography (MSCT) for assessment of ambiguous lesions is unknown. We compared both quantitative coronary angiography (QCA) and MSCT to the gold standard for a significant stenosis-minimum luminal area (MLA) by intravascular ultrasound (IVUS)-in 51 patients (64 +/- 10 years old, 19 men) with 69 angiographically ambiguous, nonleft main lesions. The MSCT was performed 17 +/- 18 days before IVUS analysis. Overall diameter stenosis by QCAwas 51.0 +/- 9.8%; 39 of 51 patients (76%) eventually underwent revascularization (38 by percutaneous coronary intervention and 1 by coronary artery bypass graft). By univariate analysis, minimum luminal diameter, MLA, lumen visibility by MSCT, and minimum luminal diameter by QCA were significant predictors of MLA by IVUS <or=4.0 mm(2). In mildly calcified lesions (calcium burden by MSCT <or=1), MLA by MSCT was a much better predictor than in more calcified lesions. By multivariate logistic regression analysis, only MLA by MSCT (odds ratio 0.754, 95% confidence interval 0.571 to 0.995, p = 0.0458) was predictive of MLA by IVUS <or=4.0 mm(2). In conclusion, in angiographically ambiguous lesions in which QCA does not distinguish significantly from nonsignificant stenosis, MSCT-measured MLA can predict significant stenosis with MLA <or=4.0 mm(2) measured by IVUS.