Background: Whether diabetes mellitus (DM) is a predictor of long-term adverse clinical outcomes after repeat drug eluting stent (DES) implantation for DES in-stent restenosis (ISR) remains controversial. We sought to evaluate the effect of DM on the long-term clinical outcomes in patients undergoing repeat DES implantation for DES-ISR lesions. Methods: In the present study, 254 patients with DES-ISR were divided into DM or non-DM groups according to the presence or absence of DM. All patients received repeat 2nd generation DES implantation for DES-ISR. The occurrences of major adverse cardiac events (MACEs) over a 2-year follow-up period were compared between the two groups. MACEs were defined as cardiac death, myocardial infarction (MI), and target lesion revascularization (TLR). MACE free survival was investigated with Kaplan-Meier curve analysis. Cox regression analysis was used to identify factors associated with MACEs. Results: Baseline clinical characteristics were similar between groups, except for the prevalence of early restenosis (lower) in the DM group. Differences in angiographic and procedural characteristics were not significant between groups. The rates of 2-year MACE (30.9 vs. 26.0%; P = 0.453) and TLR (24.7 vs. 19.7%; P = 0.411) were similar between groups. MACE-free survival and TLR-free survival were also similar between groups (P = 0.441 and P = 0.807). Subgroup analysis suggested a significant difference in the MACE (39.0 vs. 15.3%, P < 0.001) and TLR occurrence (30.5 vs. 8.2%, P < 0.001) and TLR-free survival (lower in early subgroup, P < 0.001) between early and late occurrence of ISR in the non-DM group of patients but not in the DM group. After adjustment for all significant clinical variables, Cox regression analysis indicated that DM was not associated with MACEs (hazard ratio [HR] 1.531, 95% confidence interval [CI] 0.882-2. 658, P = 0.130). Non-focal type ISR and early ISR were predictors of MACEs (HR 2.671, 95% CI 1.468-4.858, P = 0.001; HR 4. 703, 95% CI 2.725-8.117, P < 0.001, respectively). Conclusions: Patients with DM have similar 2-year clinical outcomes to patients without DM when repeat 2nd generation DES was used for treatment of DES-ISR. DM is not the predictor of long-term prognosis in patients undergoing repeat 2nd generation DES for DES-ISR.