Background: Currently it is unknown whether secondary esophagectomy after endoscopic submucosal dissection (ESD) is comparable to primary esophagectomy considering outcomes in patients with T1 esophageal cancer. We compared short- and long- term clinical outcomes between the two groups. Methods: Primary surgery (esophagectomy) was performed in 191 patients between 2003 and 2014, and 62 patients underwent secondary surgery (esophagectomy) after ESD for T1 esophageal cancer between 2007 and 2019. Propensity matching was performed for age, sex, Charlson Comorbidity Index (CCI), location, pathology, degree of differentiation, tumor size, and invasion depth. Lymph node metastasis (LNM), overall survival (OS), disease-specific survival (DSS), recurrence-free survival (RFS) and postoperative complications were compared between groups. Results: Sixty-eight patients were included after propensity score matching, LNM, OS, DSS, and RFS were comparable between the two groups. Comparing primary and secondary surgery, the respective LNM rates were 23.5% and 26.5%, 6-year OS 78.0% and 89.7%, P= 0.15; DSS were 80.4% and 96.8%, P= .057; and RFS 80.8% and 89.7%, P= .069. Multivariate analyses revealed no statistically significant differences between among LNM, OS, and RFS rates. Comparing the adverse events between the two groups, more early complications were observed in the primary surgery group than in secondary surgery group (50% vs. 20.6%, P = .021). Conclusions: Secondary surgery did not increase the risk of LNM. The long-term outcomes were comparable. Therefore, attempts to perform upfront ESD for superficial esophageal squamous cell cancers are justified.