Background: Stent underexpansion (SUE), defined by minimal stent CSA (MSA < 5 mm2), is the one of the most important predictor of in-stent restenosis (ISR) in the era of drug-eluting stent (DES) implantation. However, the difference of predictors for occurrence of ISR between SUE and non-SUE might not be well known after DES implantation. Methods: From the data in the intravascular ultrasound (IVUS) core lab at Severance Cardiovascular Hospital, Angiographically ISR was observed in 450 lesions treated with DES implantation from September 2005 to March 2009. The EXCELLENT and POET were a multicenter, randomized trial comparing paclitaxel-eluting stent (PES) (n=121), sirolimus-eluting stent (SES) (n=161), zotarolimus-eluting stent (ZES) (n=59) and everolimus-eluting stent (EES) (n=52) in patients with stable or unstable angina. A total 393 of enrolled patients who underwent post-intervention and 9-month follow-up IVUS investigation were followed clinically for one-year. We classified these into 2 groups based on the MSA; SUE group (n=107, MSA < 5 mm2) vs. non-SUE group (n=286, MSA ��� 5 mm2). Results: Significant intimal hyperplasia (IH) was defined as IH area more than 50% area of stent. Overall, 25 lesions (23.3%) among SUE group had IVUS-defined ISR, whereas 30 lesions (10.4 %) among non-SUE group has IVUS-defined ISR (p=0.005). In SUE group, diabetes mellitus [odd ratio (OR) =3.03, confidence interval (CI) =1.10-8.39, p=0.0014] and long stent (> 28mm) (OR = 3.9, CI = 1.2-12.6, p=0.017) were predictors for IVUS-defined ISR, however in non-SUE group, diabetes mellitus (OR = 2.99, CI = 1.29-6.92, p=0.014) and age (OR = 1.24, CI = 1.02-2.56, p=0.017) were predictors for IVUS-defined ISR at follow-up. Conclusions: The difference of predictors for occurrence of ISR at follow-up might exist between SUE and non-SUE group. However, in general, diabetes mellitus showed a universal predictor for occurrence of ISR at follow-up regardless of under-expansion after DES implantation.