Objectives and Background; Stent thrombosis is a serious complication of drug-eluting stent (DES) implantation. FDA panel accepts stent thrombosis but rejects increased death/MI risk for on-label DES use To date, the safety and efficacy of DES use in off-label lesion subtypes has not been systematically evaluated. We evaluated the frequency of stent thrombosis depending of on- and off-label use of DES. Method: A total 1.517 patients with DES implantation were included from March 2003 to Dec 2006 (1.560 sirolimus-eluting stents in 1.448 lesions; 715 paclitaxel-eluting stents in 634 lesions; 141 zotarolimus-eluting stents in 128 lesions). ST was newly defined by Academic Research Consortium (ARC) as reported in TCT 2006 meeting. Off-label lesion subtype was defined as follows : stent in the previously treated lesions, lesion in bypass graft, lesion length > 30 mm, reference vessel diameter (RVD) <2.5 mm or >3.5 mm, in the left main lesion, ostial lesion, bifurcation lesion, totally occluded lesion and myocardial infarction. Result; The median follow-up duration was 24 months. The frequency of off-label use of DES was 49.9% of patients. Sixty seven (4.4%) stent thrombosis occurred during median 9.6 month follow-up: 19 (1.3%) definite, 8 (0.5%) probable, 41 (2.7%) possible stent thrombosis. The incidences of stent thrombosis in off-label patients was significantly higher than the on-label patients (5.8% vs 3.0%, p=0.008). There were higher incidence of male, smoking, CRP, lower ejection fraction in the off-label patients (p<0.05). There was higher incidence of multivessel disease, complex lesion, larger reference diameter, minimal luminal diameter, longer lesion length, and total stent length in the off-label patients(p<0.05).The incidence of interruption of medication was higher in off label patients (2.7% in on-label, 5.2% in off-label, p=0.015). Other demographic, laboratory, characteristics were similar. Conclusion :The incidence of stent thrombosis was higher in the off-label patients compared with on-label patients but statistically insignificant. The large scale study was needed for the evaluation of the safety and efficacy of DES use in off-label lesion subtypes