Although previous studies have demonstrated that even quantitative coronary angiography (QCA) can not provide accurate disease morphology, there has not been a systematic comparison of disease morphology determined by QCA and intravascular ultrasound (IVUS), particularly in Japanese patients. Therefore, the present study prospectively examined patients in a multicenter cooperative study. A total of 491 coronary sites from 562 patients (446 men, 116 women; mean age, 64+/-11 years) who underwent coronary interventions were enrolled. The target lesions (>50% diameter stenosis) were evaluated pre-operatively by both QCA and IVUS operating at 30-40 MHz and the percent area stenosis, eccentricity index (EI) and lesion length were determined. The minimal (min) and maximal (max) distances from the center of the stenotic lesion to the outline of the vessel wall were measured, and the EI was calculated by the formula: [(max - min)/max]. By QCA, lesion length was determined by measuring the distance between the proximal and distal shoulders of the lesion. When the lesions were observed by IVUS with a motorized pull-back system, the length was calculated by multiplying the time for observation of the disease and 0.5 or 1 mm/s. Although the severity of the stenosis determined by QCA (86+/-10%, mean +/- SD) did not differ from that by IVUS (83+/-13%), there was no correlation between them (r=0.32, y=0.25x+65) and the correlation did not improve when lesions with remodeling, enlargement (n=176) or shrinkage (n=79) were omitted from the calculation. The EIs by QCA and IVUS were 0.51+/-0.26 and 0.52+/-0.22, respectively (NS), and there was no correlation between them (r=0.30, y=0.36x+33). However, when the lesions with remodeling were excluded, the correlation greatly improved (r=0.80, y=0.84x+10.6, p<0.05). Lesion length determined by QCA (12.4+/-6.1 mm) was significantly shorter than that by IVUS (16.3+/-8.9 mm, p<0.01). These results demonstrate that coronary angiography significantly misinterprets disease morphology in terms of severity, eccentricity and length, in part because of vessel remodeling that can be accurately determined only by IVUS.