This study was performed to investigate lesion characteristics and procedural outcomes of re-attempted percutaneous coronary intervention (PCI) for chronic total occlusion (CTO). Prior failure of percutaneous revascularization of CTO has been identified as an independent predictor of failure at subsequent attempts. However, procedural outcomes of re-attempted PCI for CTO have not been elucidated. We evaluated lesion characteristics and procedural outcomes in 310 consecutive patients undergoing CTO-PCI, and compared the results between re-attempted (n = 59) and initial procedures (n = 251). Overall, 266 CTO lesions (85.8%) were treated successfully. In addition, the technical success rate in the re-attempted CTO lesions was 69.5% (41 of 59), although this was lower than for initially attempted lesions (89.6%, 225 of 251; P = 0.0021). In the details of reasons of previous failures, treatment devices failed to cross even after guidewire cross and procedure discontinuation due to management of complications achieved higher rates of technical success compared to those with the inability of guidewire crossing in re-attempted CTO-PCI (87.5 and 85.7 vs. 65.9%, respectively). CTO lesions in which PCI was re-attempted had more complex features, including calcification, tortuous morphology, and long lesion length, whereas patient demographics were similar. Re-attempted CTO lesions required complex procedures, including the retrograde approach (55.9 vs. 13.9%, P < 0.001), longer fluoroscopic time, and larger radiation dose. Meanwhile, rates of complications and in-hospital MACCE were similarly low in both groups. The technical success rates of re-attempted CTO-PCI lesions were acceptable. Furthermore, CTO-PCIs in re-attempted lesions were as safe as initially attempted CTO-PCI. However, re-attempted CTO-PCI lesions showed features of high anatomical complexity that required more complex and longer procedures, including the retrograde approach, for successful interventional revascularization. Re-attempted CTO-PCI due to treatment devices failed to cross even after guidewire cross and procedure discontinuation due to management of complications in previous attempt had higher success rates that those with the inability of guidewire crossing.