Abstract Background Proton beam therapy (PBT) for esophageal squamous cell carcinoma (ESCC) is a promising curable treatment to avoid surgery, however the long-term outcomes have not enough been reported. Herein, we compared the survival outcomes of PBT with those of esophagectomy after neoadjuvant chemotherapy, which is the standard of care in Japan. Methods Patients with thoracic ESCC, clinical stage II or III, who underwent neoadjuvant chemotherapy plus radical resection and proton beam therapy between 2011 and 2020 were identified from database of our institute. The chemotherapy regimen was 5FU, CDDP (CF) with/without docetaxel. PBT was defined as irradiation of 50 GyE or more to the primary tumor, prophylactic irradiation, and concurrent chemotherapy (CF). Patients with other severe diseases and patients who could not be followed up were excluded. To adjust confounding factors, propensity score matching was performed; covariates included T and N-factor, age, gender, time of treatment and site of tumor. We retrospectively evaluated overall survival time of the surgery and PBT groups as primary outcome in this study. Results 244 patients were selected from the database. Median age was 71 years, 209 were male, 102 had esophagectomy after chemotherapy, and 142 had PBT. After propensity score matching, 46 patients in each group were selected by propensity score matching. Median survival was 42.8:44.6 months in the surgery and PBT groups, respectively (p=0.784). There were no treatment-related deaths, and Grade 3 or higher adverse events occurred in 8 and 8 patients (p=1.00), respectively. Conclusion PBT for stages II and III ESCC was comparable to neoadjuvant chemotherapy followed by surgery in long-term outcomes and adverse events.