Background Although drug-eluting stents (DES) has shown to be effective in angiographic restenosis, percutaneous coronary intervention (PCI) of long coronary lesions are still challenging and limited for the long-term experience in the current era. Methods We performed subgroup analysis of patients with very long coronary lesions (lesion length ��� 28 mm) in the ZEST trial comparing second-generation zotarolimus-eluting stents with first-generation sirolimus- and paclitaxel-eluting stents. The primary end point was composite major adverse cardiac events (MACE) of death, myocardial infarction (MI), or ischemic-driven target vessel revascularization (TVR). Results A total of 960 patients with very long coronary lesions were treated with zotarolimus- (ZES, 322 patients), sirolimus- (SES, 317 patients), or paclitaxel-eluting stents (PES, 321 patients). All patients with very long lesions were available for 2-year clinical follow-up. At 24 months, the zotarolimus-stent group showed similar rates of MACE compared with the paclitaxel-stent group (16.1% vs. 17.4, P=0.58), but trend toward higher rates of events compared with the sirolimus-stent group (16.0% vs. 11.0%, p=0.059). Theses difference were mainly driven by TVR (SES vs. ZES, p=0.059; ZES vs. PES, p=0.37; SES vs. PES, p=0.006). There was no significant difference of death or MI among the groups. Conclusion For patients with very long coronary artery disease, PCI with ZES results in similar rates of MACE compared with PES, but in trend toward higher rates of MACE compared with SES at 2-year.