Background The prognosis of elderly patients with chronic total occlusion (CTO) and diabetes mellitus (DM) treated with percutaneous coronary intervention (PCI) is not known. Objective To investigate the effect of diabetes on long-term follow-up of CTO after PCI in elderly patients. Methods A total of 153 elderly patients (age > 65 years old) with CTO lesions which were successfully treated with PCI were enrolled. Fifty one patients with diabetes and 102 without diabetes were compared for long-term outcomes (mean follow up: 36 +/- 12 months). Major adverse cardiac events (MACE) which include death, myocardial infarction or target lesion revascularization (TLR) were considered as a combined endpoint. Results The combined endpoint occurred in 29.4% of diabetes patients, and 11.3% of the patients without diabetes (P < 0.05). The Cox proportional hazards model identified: drug eluting stent (DES) or bare metal stent (BMS) (HR: 0.13, 95% confidence interval (95% CI): 0.03-0.62, P = 0.004), DM (HR: 6.69, 95% CI: 1.62-15.81, P = 0.01) and final minimal lumen diameter (MLD) (HR: 0.37, 95% CI: 0.13-0.90, P = 0.03) as independent predictors of MACE, DM with renal impairment (HR: 6.64, 95% CI: 1.32-33.36, P = 0.02), HBA1C on admission (HR: 1.79, 95% CI: 1.09-2.94, P = 0.02), as independent predictors of MACE at long term follow-up. Conclusions The study demonstrates that DM is a predictive factor for MACE in elderly CTO patients treated with PCI, type of stent, final minimal lumen diameter and DM with renal impairment, and HBA1C level on admission are predictors of MACE. J Geriatr Cardiol 2013; 10: 16-20. doi: 10.3724/SP.J.1263.2012.11091