Background : After coronary stenting, the restenosis rate is known to be higher in DM than non-DM patients due to excessive intimal hyperplasia. Rosiglitazone, a insulin-sensitizing agent, has anti-inflammatory property acting on monocytes, smooth muscle cells, endothelial cells, etc. as a PPAR-款 agonist. We investigated the efficacy of rosiglitazone for preventing restenosis after coronary stenting in diabetic patients. Methods : We conducted a prospective case-controlled trial involving 95 diabetic patients (control n=48, rosiglitazone (4 mg per day) group n=47) undergoing coronary stenting. Serum glucose and cholesterol levels were tried to be optimized using additional hypoglycemic agents and statins. The primary end point was the binary restenosis rate at 6-month angiographic follow-up. Results : A follow-up angiography was performed in 83 patients including 45 patients with 55 lesions in control group (m:f = 31:11, age 59.9짹9.3 years), and 38 patients with 51 lesions in the rosiglitazone group (m:f=24:18, age 60.9짹9.3 years). The baseline clinical profile and blood chemistry between two groups were not different. Baseline angiographic data of both groups were as following : reference diameter 3.15짹0.49 vs. 3.16짹0.49 mm (p=NS), minimal lumen diameter (MLD) 0.65짹0.41 vs. 0.83짹0.57 mm (p=NS), diameter stenosis (DS) 79.4짹12.8 % vs. 74.4짹15.8 %, lesion length 16.48짹5.16 vs. 19.02짹6.09 mm (p<0.05). There was no difference in MLD or acute gain after coronary stenting between the two groups. At 6-month follow-up, the rosiglitazone group showed a significantly lower restenosis rate than the control group (17.6% vs. 38.2%, p = 0.03). The follow-up DS (24.2 짹 23.2% vs. 42.9 짹 32.2%, p = 0.001) and loss index (0.49짹0.41 vs. 0.27짹0.30, p=0.008) were also lower in the rosiglitazone group. Major adverse cardiac events occurred less frequently in the rosiglitazone group (10.5% vs 20%, p=0.244) at 6 months, although it was not statistically significant. Conclusions : Rosiglitazone effectively reduced restenosis in diabetic patients after coronary stenting. This result is probably related to pleotrophic property of rosiglitazone rather than its hypoglycemic effect.