The close proximity of esophagus to the left atrial posterior wall predisposes esophagus to thermal injury during catheter ablation for atrial fibrillation (AF). In this retrospective study, we aimed to investigate risk factors of esophageal injury (EI) caused by catheter ablation for AF. Patients who underwent first-time AF ablation from July 2013 to June 2018 were included. The esophagus was visualized by oral soluble contrast during ablation for all patients and a subset of patients were selected to undergo endoscopic ultrasonography (EUS) to estimate EI post ablation. Degree of EI was categorized as Kansas City classification: type 1: erythema; type 2: ulcers (2a: superficial ulcers; 2b: deep ulcers); type 3: perforation (3a: perforation without communication with the atria; 3b: atrioesophageal fistula [AEF]). Of 3,852 patients, 236 patients (61.5 ± 9.7 years; male, 69%) received EUS (EUS group) and 3616 (63.2 ± 10.9 years; male, 61.1%) without EUS (No-EUS group). In EUS group, EI occurred in 63 patients (type 1 EI in 35 and type 2 EI in 28), and no type 3 EI was observed during follow up. In a multivariable logistic regression analysis, an overlap between the ablation lesion and esophagus was an independent predictor of EI (odds ratio, 21.2; 95% CI: 6.23-72.0; P < 0.001). In No-EUS group, esophagopericardial fistula (EPF; n = 3,0.08%) or AEF (n = 2,0.06%) was diagnosed 4-37 days after ablation. In 3 EPF patients, 2 completely recovered with conservative management and 1 died. Two AEF patients died. Ablation at the vicinity of the esophagus predicts risk of EI. EUS post ablation may prevent the progression of EI and should be considered in management of EI. It remains challenging to identify patients with high risk of EI. Pulmonary vein isolation (PVI) has emerged as a cornerstone therapy for atrial fibrillation (AF) ablation. The close proximity of esophagus to the left atrial posterior wall predisposes esophagus to thermal injury during catheter ablation for AF and Esophageal perforation to the left atrium is a fatal complication 1-5. Introduction of contact force-sensing catheters and cryo-balloon catheters appears to make no change in procedural complication rates for patients undergoing AF ablation 6 , and the utility of contact force-sensing (CF-sensing) catheter may be associated with increased rates of atrioesophageal fistula (AEF) formation 7. A recent study appreciably suggested that postprocedural gastroesophageal endoscopy (GSE) could identify EI, and higher intraesophageal temperature measured by a luminal esophageal temperature probe (LET) was associated with the occurrence of EI 5. However, the protective effect of LET remains controversial since LET itself may also cause EI 8,9 , and routine GSE post AF ablation was not available. Until now, there is no widely accepted approach to minimize esophageal injury (EI) caused by catheter ablation for AF. We sought to identify risk factors of EI caused by ablation for AF, and explore the potential management of EI in this study.