The intravascular ultrasound (IVUS) criteria for stent optimization have not been determined in stenting long lesions. We evaluated the predictors of angiographic restenosis and compared it with stent lumen cross-sectional area (CSA) and stent length between short (stent length <20 mm) and long (> or =20 mm) coronary stenting. IVUS-guided coronary stenting was successfully performed in 285 consecutive patients with 304 native coronary lesions. Six-month follow-up angiogram was performed in 236 patients (82.8%) with 246 lesions (80.9%). Results were evaluated using conventional (clinical, angiographic, and IVUS) methods. The overall angiographic restenosis rate was 22.8% (56 of 246 lesions) (short stent 17.6% vs. long stent 32.2%, p = 0.009). Using multivariate logistic regression analysis, the independent predictors of angiographic restenosis were the IVUS stent lumen CSA (odds ratio 1.51, 95% confidence intervals 1.18 to 1.92, p = 0.001) and stent length (odds ratio 0.95, 95% confidence intervals 0.91 to 1.00, p = 0.039). The angiographic restenosis rate was 54.8% for stent lumen CSA of <5.0 mm2 (short stent 37.5% vs. long stent 73.3%, p = 0.049), 27.4% for CSA between 5.0 and 7.0 mm2 (short stent 24.1% vs. long stent 31.7%, p = 0.409), 10.5% for CSA between 7.0 and 9.0 mm2 (short stent 10.0% vs. long stent 12.5%, p = 0.772), and 11.4% for stent lumen CSA of > or =9.0 mm2 (short stent 10.4% vs. long stent 13.3%, p = 0.767) (p = 0.001). Compared with short coronary stenting, long coronary stenting is effective treatment modality to cover long lesions with comparable long-term clinical outcomes in cases of stent lumen CSA of > or =7.0 mm2. Regardless of the stent length, the most important factor determining angiographic restenosis was the IVUS stent lumen CSA in relatively large coronary artery lesions.