Background: There is a little data about chronic total occlusive in-stent restenosis (CTO-ISR). So we examined clinical outcomes of percutaneous coronary intervention for CTO-ISR lesion. Methods: From January 1999 to May 2011, we collected 56 patients with CTO-ISR lesion (DES CTO-ISR, n=30 and BMS CTO-ISR, n=26), and compared their clinical and angiographic data between the BMS CTO-ISR group and DES CTO-ISR group. retrospectively. Result: The majority of the patients were men (68%) with a mean age of 60.4짹10.4. The CTO-ISR vessels were predominantly the left anterior descending artery (46%) and 43(77%) of the patients had multivessel disease. In baseline characteristics, there were no significant differences between both groups except for the duration from index PCI to PCI for CTO-ISR (BMS; 1602.7짹1210.0 days vs DES; 882.2짹726.2 days, p=0.03). Overall procedural success rate for ISR-CTO was 73% (41/56 PCIs). The success rate of BMS ISR-CTO was 62% (16/26 PCIs) and DES CTO-ISR was 83% (25/30 PCIs). DES CTO-ISR group was significant higher success rate than BMS CTO-ISR (OR=0.24; 95% CI 0.064-0.929, p=0.04). The duration from index PCI to CTO-ISR manifestation, stent length, kinds of previous stent (BMS vs. DES) were independent predictor for PCI success. Conclusion: Our data showed higher success rate in PCI for DES CTO-ISR than the BMS CTO-ISR and the duration from index PCI to CTO-ISR manifestation was different between the both groups. These suggest that a mechanism forming CTO-ISR between the both groups might be different. Further study is needed for the optimal PCI strategy for CTO-ISR lesion.