BACKGROUND: The safety of drug-eluting stents has been called into question by recent reports of increased stent thrombosis, myocardial infarction, and death. Such studies have been inconclusive because of their insufficient size, the use of historical controls, a limited duration of follow-up, and a lack of access to original source data. METHODS: We performed a pooled analysis of data from four double-blind trials in which 1748 patients were randomly assigned to receive either sirolimus-eluting stents or bare-metal stents and five double-blind trials in which 3513 patients were randomly assigned to receive either paclitaxel-eluting stents or bare-metal stents; we then analyzed the major clinical end points of the trials. RESULTS: The 4-year rates of stent thrombosis were 1.2% in the sirolimus-stent group versus 0.6% in the bare-metal-stent group (P=0.20) and 1.3% in the paclitaxel-stent group versus 0.9% in the bare-metal-stent group (P=0.30). However, after 1 year, there were five episodes of stent thrombosis in patients with sirolimus-eluting stents versus none in patients with bare-metal stents (P=0.025) and nine episodes in patients with paclitaxel-eluting stents versus two in patients with bare-metal stents (P=0.028). The 4-year rates of target-lesion revascularization were markedly reduced in both the sirolimus-stent group and the paclitaxel-stent group, as compared with the bare-metal-stent groups. The rates of death or myocardial infarction did not differ significantly between the groups with drug-eluting stents and those with bare-metal stents. CONCLUSIONS: Stent thrombosis after 1 year was more common with both sirolimus-eluting stents and paclitaxel-eluting stents than with bare-metal stents. Both drug-eluting stents were associated with a marked reduction in target-lesion revascularization. There were no significant differences in the cumulative rates of death or myocardial infarction at 4 years.