Randomized clinical trials have shown that sirolimus-eluting stents (SESs) decrease restenosis rates compared with bare metal stents (BMSs), but their efficacy among patients who have diabetes mellitus remains to be established. This study investigated the effect of SES implantation in a high-risk population (i.e., patients who had diabetes and small coronary vessel disease). For this purpose, we analyzed outcomes of the subset of patients who had diabetes and were enrolled in the SES-SMART, a randomized trial that compared the results of implantation of SESs and BMSs in small coronary arteries. Twenty-nine patients who had diabetes were originally randomized to receive SESs and 45 patients received BMSs. The use of SESs was associated with approximately 60% decreases in the relative incidence of in-segment angiographic restenosis (63% vs 25%, p = 0.003) and in-segment late loss (0.76 vs 0.28 mm, p <0.002). Angiographic patterns of restenosis were more favorable in the SES group. SES implantation was associated with a 15% absolute decrease in adverse clinical events. In patients who had insulin-dependent diabetes mellitus, SESs showed a high in-segment restenosis rate (40%) that was principally due to persistent restenosis. In conclusion, in diabetics with small coronary arteries, SES implantation significantly reduces the incidence of the 8-month angiographic restenosis rate compared with BMSs.