The risk of late thrombotic events and the need for prolonged dual antiplatelet therapy detract from the clinical advantage offered by drug-eluting stents (DESs). Short-term studies have shown premature clopidogrel cessation to be a strong predictor of stent thrombosis (ST) after DES implantation. Data pertaining to the utility of clopidogrel therapy and its optimal duration to prevent late ST remain limited. The study population consisted of 2,889 patients who underwent unrestricted intracoronary DES implantation from April 2003 to January 2007 for whom clopidogrel compliance data were available. Definite ST proved by angiography or autopsy within 12 months of the index procedure occurred in 61 patients. Comparisons of clinical and procedural characteristics in addition to outcomes (death and Q-wave myocardial infarction) were made between the ST and no-ST (2,828 patients) groups. Clopidogrel compliance was assessed at all follow-up time points. For patients in the ST group, clopidogrel compliance status for the remaining study period was defined as that at the time of ST. Logistic regression analysis was performed at 30 days, 6 months, and 12 months to identify independent predictors of cumulative ST. Patients with ST were more likely to have previous congestive heart failure and worse left ventricular ejection fraction. ST was associated with significantly higher mortality at 12 months (23.5% vs 3.2%, p <0.001). Clopidogrel compliance was 80.2% in the overall population and 73.8% in patients presenting with ST (82.6% in patients presenting with early ST and 43.8% in those with late ST). By logistic regression analysis, clopidogrel cessation was an independent predictor of cumulative ST at 30 days and 6 months but not at 12 months. In conclusion, high rates of clopidogrel compliance can be achieved in contemporary practice. Clopidogrel cessation by 12 months is no longer predictive of ST, thus suggesting the optimal duration of therapy for the prevention of ST to be 6 to 12 months.