Background: Stent thrombosis (ST) is still a major limitation of drug-eluting stent (DES) usage, whether DES increase the risk of ST than bare metal stent, or not. The prediction of ST in acute myocardial infarction (AMI) with the existing variables is poor. The aim of the present study is to include the clinical and angiographic variables associated with ST and develop a prediction model for ST from the Korean Acute Myocardial Infarction (KAMI) Registry database. Methods: From a total cohort of KAMI registry (6500 patients) since January 2007 to December 2007, 1807 patients (27.8%) were available to analyze this study. ST was defined according to ARC definition and divided as early ST (0-30 days) and late ST (> 30 days). Results: Clinical follow-up over 1 month was available in 88%, more than 12 months in 55.4%. Overall incidence of ST was 2.2% (39 patients) in KAMI registry. Early ST was developed in 1.3%, late ST in 0.9%. Male gender, relatively younger age, advanced Killip classification, total number of implanted stent, aspiration before stent procedure were univariate predictors of ST. Advanced Killip classification (hazard ratio 2.37, 95% CI 1.13-5.37, p=0.023), total number of implanted stent (1.66, 1.17-2.37, p=0.005) were independent predictors of ST. Conclusion: The incidence of ST after stent implantation in AMI (2.2%) was relatively higher than previously reported clinical trials for stable angina. According to the cohort of KAMI registry, advanced Killip classification and total number of implanted stent were independent predictors of ST in KAMI registry.