BACKGROUND: Drug-eluting stents are highly effective in reducing the rate of in-stent restenosis. It is not known whether there are differences in the effectiveness of currently approved drug-eluting stents in the high-risk subgroup of patients with diabetes mellitus. METHODS: We enrolled 250 patients with diabetes and coronary artery disease: 125 were randomly assigned to receive paclitaxel-eluting stents, and 125 to receive sirolimus-eluting stents. The primary end point was in-segment late luminal loss. Secondary end points were angiographic restenosis (defined as in-segment stenosis of at least 50 percent at follow-up angiography) and the need for revascularization of the target lesion during a nine-month follow-up period. The study was designed to show noninferiority of the paclitaxel stent as compared with the sirolimus stent, defined as a difference in the extent of in-segment late luminal loss of no more than 0.16 mm. RESULTS: The extent of in-segment late luminal loss was 0.24 mm (95 percent confidence interval, 0.09 to 0.39) greater in the paclitaxel-stent group than in the sirolimus-stent group (P=0.002). In-segment restenosis was identified on follow-up angiography in 16.5 percent of the patients in the paclitaxel-stent group and 6.9 percent of the patients in the sirolimus-stent group (P=0.03). Target-lesion revascularization was performed in 12.0 percent of the patients in the paclitaxel-stent group and 6.4 percent of the patients in the sirolimus-stent group (P=0.13). CONCLUSIONS: In patients with diabetes mellitus and coronary artery disease, use of the sirolimus-eluting stent is associated with a decrease in the extent of late luminal loss, as compared with use of the paclitaxel-eluting stent, suggesting a reduced risk of restenosis. Copyright 2005 Massachusetts Medical Society.