Background: To compare radical surgery with definitive radiotherapy for upper third esophageal squamous cell carcinoma. Materials and Methods: A total of 191 patients were included in the study. Patients’ clinicopathologic features, and survival time were recorded. Kaplan-Meier (K-M) analysis was adopted to analyze Overall survival (OS), Disease-free survival(DFS), Progression-free survival(PFS), and a Cox multivariate model was used to adjust potential confound factors. Results: The K-M survival analysis showed that treatments, location of lesion, and length of lesion were all associated with the OS (P<0.005). In the surgery group, K-M survival  showed that T stage (T1 vs T2, P=0.012, T1 vs T3, P=0.002), location (upper vs upper merged middle, P<0.001), and length lesion (<5 cm vs >5 cm, P=0.015), affected the OS，T stage (T1 vs T2, P=0.018, T1 vs T3, P=0.020) and location of lesion (upper vs upper merged middle, P=0.007) was associated with DFS. The Cox model showed that none of these parameters independently influenced the OS and DFS. In the radiotherapy group, K-M survival showed that supraclavicular lymph node metastasis (P=0.007), concurrent chemo-radiotherapy (P=0.012), and sex (P=0.047) influenced the OS，adjuvant chemotherapy (P=0.013) and age (P=0.013) influenced PFS, The Cox model showed that supraclavicular lymph node metastasis (P=0.018) independently influenced OS and adjuvant chemotherapy (P=0.046) independently influenced PFS. Conclusion: Surgery has better therapeutic effect than radiotherapy. Patients with an upper merged middle lesion and advanced T stage for surgery, male, local advanced and without concurrent chemo-radiotherapy for radiotherapy have a poor prognosis.