Aims To investigate whether intravascular ultrasound provides additional information regarding the prediction of stent thrombosis, a retrospective multicentre registry was designed to enrol patients with stent thrombosis following stent deployment under ultrasound guidance. Methods and Results A total of 53 patients were enrolled (mean age 61±9 years) with stable angina (43%), unstable angina (36%), and post-infarct angina (21%) who underwent intracoronary stenting. The majority had balloon angioplasty alone prior to stenting (94%) with 6% also undergoing rotational atherectomy. The indication for stenting was elective (53%), suboptimal result (32%) and bailout (15%). There were 1·6±0·8 stents/artery with 87% undergoing high-pressure dilatation (≥14 atmospheres). The minimum stent area was 7·7±2·8mm2with a mean stent expansion of 81·5±21·9%. Overall, 94% of cases demonstrated one abnormal ultrasound finding (stent under-expansion, malapposition, inflow/outflow disease, dissection, or thrombus). Angiography demonstrated an abnormality in only 32% of cases (chi-square=30·0,P <0·001). Stent thrombosis occurred at 132±125h after deployment. Myocardial infarction occurred in 67% and there was an overall mortality of 15%. Conclusion On comparison with angiography, the vast majority of stents associated with subsequent thrombosis have at least one abnormal feature by intravascular ultrasound at the time of stent deployment.