Background— To expand the paucity of data on the efficacy of various drug-eluting stents in diabetic patients. Methods and Results— Type 2 diabetic patients treated in our institution from October 2005 to January 2007 presenting with of de novo lesions in native coronary arteries were randomly assigned to sirolimus-eluting stents (Cypher group; n=76); paclitaxel-eluting stents (Taxus group; n=75); and everolimus-eluting stents (Endeavor group; n=75). Poor metabolic control (HbA1c >7% and low-density lipoprotein cholesterol >100 mg/dL) and microvascular complications (retinopathy and/or nephropathy) were assessed. The primary end point was the 3-year composite of major adverse cardiac events (MACE), including death of any cause, myocardial infarction, and clinically driven target vessel revascularization. MACE-free survival was 86.8% in the Cypher group, 82.5% in the Taxus group, and 64.4% in the Endeavor group ( P =0.006 by log-rank test). The post hoc comparisons showed no significant difference between Cypher versus Taxus groups (adjusted P =1.0) but a higher MACE rate in the Endeavor group versus both the Cypher group (adjusted P =0.012) and the Taxus group (adjusted P =0.075). Independent predictors of 3-year MACE at Cox analysis were treatment by Endeavor versus Cypher stent (2.35 [95% confidence interval, 1.07 to 5.41]; P =0.030), multivessel disease (hazard ratio, 1.78 [95% confidence interval, 1.06 to 2.66]; P =0.031), diabetic retinopathy (hazard ratio, 1.60; [95% confidence interval, 1.03 to 2.76]; P =0.038), and poor metabolic control (hazard ratio, 1.60; [95% confidence interval, 1.02 to 2.52]; P =0.048). Conclusions— The present pilot study suggests that in diabetic patients, the Endeavor stent is associated with a higher 3-year MACE rate when compared with Cypher and Taxus stents.