Diabetes mellitus is a strong predictor of in-stent restenosis after bare-metal stent placement because of exaggerated intimal hyperplasia (IH). In this study, serial intravascular ultrasound was used to evaluate the effects of polymer-based, paclitaxel-eluting stents (PES) on in-stent neointima formation in patients with and without diabetes. A volumetric IVUS meta-analysis of 956 patients randomized to PES or bare-metal stents in the TAXUS IV (n=268), TAXUS V (n=509), and TAXUS VI (n=179) trials was performed. Diabetes was present in 273 of the 956 patients (28.6%). Stent, luminal, and IH areas were measured every millimeter within the stent, and volumes, percentage IH (%IH), and neointima-free stent length were calculated. Patients with diabetes treated with bare-metal stents had greater IH volumes (78.9 vs 61.2 mm3, p=0.0095) and %IH (34.9% vs 30.1%, p=0.0186) compared with those without diabetes. However, patients with diabetes treated with PES had similar median IH volume (22.8 vs 16.3 mm3, p=0.35) and median %IH (9.1% vs 9.2%, p=0.27) compared with those without diabetes and significantly less IH volume and %IH than after bare-metal stent implantation (p<0.0001 for the 2 comparisons). In conclusion, serial volumetric IVUS analysis showed that treatment with PES neutralizes the propensity of patients with diabetes to develop greater IH, such that patients with and without diabetes treated with PES have similar vascular responses and restenosis rates.