Objective To assess the distribution of pressure and shear-related forces acting on atherosclerotic plaques and their association with lesion characteristics using coronary CT angiography (cCTA)-based computational fluid dynamics (CFD) model of epicardial coronary arteries. Methods Patient-specific models of epicardial coronary arteries were reconstructed from cCTA in 80 patients (12 women, 63.89.0years). The pressure and wall shear stress (WSS) in left anterior descending coronary arteries were assessed using CFD. High-risk plaques were defined as the presence of at least one of the following adverse plaque characteristics: low-density plaque, positive remodelling, napkin-ring sign and spotty calcification. Results At resting condition, 39.5% of stenotic segments (% diameter stenosis 52.3 +/- 14.4%) were exposed to high WSS (>40dyne/cm(2)). When the stenotic lesion was subdivided into three segments, the distribution of WSS was different from that of pressure change and its magnitude was highest at minimal lumen area (p<0.001). High pressure gradient, proximal location, small lumen and short length were independent determinants of WSS (all p<0.05). The plaques exposed to the highest WSS tertile had a significantly greater proportion of high-risk plaques. The addition of WSS to % diameter stenosis significantly improved the measures of discrimination and reclassification of high-risk plaques (area under the curves from 0.540 to 0.718, p=0.031; net reclassification index 0.827, p<0.001). Conclusions The cCTA-based CFD method can improve the identification of high-risk plaques and the risk stratification for coronary artery disease patients by providing non-invasive measurements of WSS affecting coronary plaques.