The feasibility of expanding the indications for endoscopic submucosal dissection to treat early gastric cancer based on long-term outcomes has shown conflicting results. This study aimed to investigate whether outcomes or adverse events associated with endoscopic submucosal dissection are comparable to those of surgery for early gastric cancer that including the absolute and expanded indications.Data of 159 early gastric cancers from 153 patients treated with endoscopic submucosal dissection or surgery between January 2004 and October 2014 were reviewed retrospectively. Early gastric cancers fulfilled the absolute or expanded indications with differentiated type adenocarcinoma were included.The endoscopic submucosal dissection and surgery group showed no significant difference in the incidence of residual disease (P = .48), local recurrence (P = .46), and metachronous cancer (P = .22). Kaplan-Meier analysis showed no significant difference in 2-year (97.6% versus [vs] 92.4%; P = .45) and 5-year (95.8% vs 95.6%; P = .26) overall survival rate between 2 groups. There was also no significant difference in 2-year (100% vs 94.1%; P = .98) and 5-year (100% vs 98.4%; P = .89) disease-free survival rate. Early and late adverse events also showed no significant differences.For the treatment of early gastric cancer fulfilled absolute and expanded indications, endoscopic submucosal dissection is not inferior modality regarding the clinical outcomes and safety, compared with surgery.