Although the safety of drug-eluting stents has been under considerable scrutiny, limited real-world follow-up data extending up to 4 years are available. The randomized clinical trials carefully selected patients and are not reflective of everyday practice. From April to October 2002, 508 consecutive patients treated with sirolimus-eluting stents (SES) were enrolled. The control group consisted of 450 patients treated with bare-metal stents during the preceding 6 months. After 4 years of follow-up, the incidence of composite major adverse clinical events (all-cause death, myocardial infarction, or target vessel revascularization) was found to be significantly lower in the SES group (23.0% vs 28.7%, adjusted hazard ratio 0.66, 95% confidence interval 0.51 to 0.86), as were rates of target vessel revascularization (12.2% vs 17.8%, adjusted hazard ratio 0.57, 95% confidence interval 0.39 to 0.83). There were no differences in all-cause mortality (10.5% for SES vs 10.6% for bare-metal stents, p = 0.9) or in the rates of cardiac death (4.5% vs 6.9%, p = 0.1). Although there was no difference in overall stent thrombosis (2.3% vs 2.2%, p = 1.0), SES had a higher rate of very late stent thrombosis (1.4% vs 0%, p = 0.02), balanced by a lower rate of early stent thrombosis (0.4% vs 1.8%, p = 0.05). In conclusion, after 4 years, SES were found to remain safe and effective compared with bare-metal stents. Nevertheless, the higher rate of very late stent thrombosis remains a concern. Longer term follow-up will be required to determine the extent of this problem.