BACKGROUND: Features of ruptured and high-risk plaque have been described on coronary computed tomography angiography (coronary CTA), but not systematically assessed against intravascular ultrasound (IVUS). We examined the ability of coronary CTA to identify IVUS defined ruptured plaque and Virtual Histology Intravascular Ultrasound (VH-IVUS) defined thin-cap fibroatheroma (TCFA). METHODS: Sixty-three patients (32 with acute coronary syndrome and 31 with stable angina) underwent coronary CTA, IVUS and VH-IVUS. Plaque rupture on CTA was defined as intra-plaque contrast and its frequency compared with IVUS-defined plaque rupture. We then examined the relationship of conventional coronary CTA high-risk features (low attenuation plaque, positive remodeling, spotty calcification and the Napkin-Ring sign) in VH-IVUS-defined TCFA. We compared these with a novel index based on quantifying the ratio of necrotic core to fibrous plaque using x-ray attenuation cut-offs derived from the relationship of plaque to luminal contrast attenuation. RESULTS: Of the 71 plaques interrogated with IVUS, 39 were ruptured. Coronary CTA correctly detected 13-ruptured plaques with 3 false positives giving high specificity (91%) but low sensitivity (33%). None of the conventional coronary CTA high-risk features were significantly more frequent in the higher-risk (VH-IVUS defined thin-cap) compared with thick-cap fibroatheroma. However, the new index (necrotic core/fibrous plaque ratio) was higher in thin-cap (mean 0.90) vs. thick-cap fibroatheroma (mean 0.59), p < 0.05. CONCLUSIONS: Compared with intravascular ultrasound, coronary CTA identifies ruptured plaque with good specificity but poor sensitivity. We have identified a novel high-risk feature on coronary CTA (necrotic core/fibrous plaque ratio that is associated with VH-IVUS defined-TCFA.