The present trial was conducted to evaluate efficacy and safety of sirolimus-eluting stent (SES), Cypher compared with paclitaxel-eluting stent (PES), Taxus among patients with de novo coronary lesions. Patients were randomized to stent implantation with either Taxus (669 patients with 941 lesions) or Cypher (684 patients with 970 lesions). Direct stenting was allowed. Use of aspirin and clopidogrel and glycoprotein IIb/IIIa inhibitor was at the physician's discretion. There were 1911 lesions in the 1353 patients with an average of 1.94 stents used per patient. Baseline clinical and angiographic characteristics were wellmatched between the treatment groups, with 28% diabetics. Lesions were relatively complex, with 86.4% classified as BII or C, 27.6% >20 mm in length, and 5.2% bifurcation lesions. Procedural success was 95% in each group. At 8 months angiographic follow-up, Cypher group had larger in-stent minimum lumen diameter (2.00 mm v. 1.85 mm), less late in-stent lumen loss (0.09 mm v. 0.31 mm), and reduced percent diameter stenosis (23.1% v. 26.7%), (p<0.0001 for all comparisons). However, the primary endpoint of binary in-lesion restenosis did not differ between the treatment groups (9.6% for Cypher group v. 11.1% for Taxus stent group, p=0.32). At 8 months clinical follow-up, there was no difference in secondary endpoint of major adverse cardiac event (MACE) (9.2% for Cypher group v. 10.6% for Taxus group, p=0.41) or any component of MACE [death 1.8% v. 1.2%, p=0.50; myocardial infarction (MI) 4.8% v. 5.5%, p=0.62; target lesion revascularization 5.0% v. 5.4%, p=0.81, respectively]. Stent thrombosis by 30 days was higher in the Taxus stent group (1.8% v. 0.4%, p=0.0196 for astreated analysis; 1.6% v. 0.6%, p=0.0723 for intent-to-treat analysis). In conclusion, among patients with de novo coronary lesions, despite better angiographic parameters at 8 months i.e. minimum lumen diameter, late lumen loss, and percent diameter stenosis in the Cypher group, there was no difference in the primary endpoint of binary restenosis or in the clinical MACE rate at 8 months, suggesting both drug-eluting stents are effective in reducing restenosis. Although the absolute number of stent thrombosis was small, the increase in-stent thrombosis through 30 days in Taxus group needs close observation.