This study sought to comprehensively identify predictors of stent thrombosis (ST). Given the devastating consequences of ST, efforts should be directed toward risk stratification to identify patients at highest risk for ST. Consecutive patients with angiographic ST were enrolled. Patients who did not suffer from a ST were randomly selected in a 2:1 ratio and were matched for: 1) percutaneous coronary intervention (PCI) indication; 2) same date of index PCI; and 3) same interventional center. Of 21,009 patients treated with either a bare-metal or drug-eluting stent, 437 patients (2.1%) presented with a definite ST. A total of 140 STs were acute, 180 were subacute, 58 were late, and 59 were very late. Undersizing of the coronary stent, Thrombolysis In Myocardial Infarction flow grade <3, present malignancy, presence of intermediate coronary artery disease proximal and distal to the culprit lesion, dissection, lack of aspirin, bifurcation lesions, ejection fraction <30%, and younger age were associated with ST. The lack of clopidogrel therapy at the time of ST in the first 30 days after the index PCI (hazard ratio HR: 36.5, 95% confidence interval CI: 8.0 to 167.8), between 30 days and 6 months after the index PCI (HR: 4.6, 95% CI: 1.4 to 15.3), and beyond 6 months (HR: 5.9, 95% CI: 1.7 to 19.8) after the index PCI was strongly associated with ST. Important correlates of ST were identified. Discontinuation of clopidogrel, undersizing of the coronary stent, present malignancy, and intermediate (>or=50% to <70% stenosis) coronary artery disease proximal to the culprit lesion were the strongest predictors of ST.