BACKGROUND: The matter of postoperative radiotherapy (PORT) in esophageal cancer (ESCA) was far from conclusive. Some evidence indicated that lymph node status could affect treatment. We evaluated lymph node ratio (LNR) as an indicator that could be applied to predict PORT benefit. METHODS: Retrospective cohort study collected the data of N1, N2, N3 stage ESCA patients from the Surveillance, Epidemiology, and End Results database (SEER) to analyze the association between LNR and prognosis from 2004 to 2015. Patients were categorized into two subsets based on the LNR cut-off value of 0.23 using receiver operating characteristic curve (ROC). Kaplan-Meier analysis was utilized to estimate the proportion of overall survival (OS) and esophagus cancer-specific survival (CSS) in two LNR groups. Cox regression analysis and competitive risk model was adopted to investigate the impacts of LNR on prognosis. RESULTS: Of 2,165 ESCA patients identified, 1,165 (53.8%) had LNR>0.23. The LNR was an independent prognostic factor and associated with better OS and CSS of LNR≤0.23 (P < 0.001). In competitive risk model, a worse CSS was analyzed of LNR>0.23 (HR = 1.71; 95% CI 1.53-1.91). Subgroup analyses indicated that PORT was associated with favorable OS and CSS. Furthermore, when stratified by Node stage, PORT was associated with a survival benefit only in N1 stage with higher LNR (LNR>0.23) after adjusting for other covariates. CONCLUSIONS: LNR exceeding 0.23 was negatively associated with prognosis in ESCA. The survival benefit from PORT in ESCA seems to be limited to LNR of 23% or more only in N1 stage. This study highlights the biomarker meaning of LNR on identifying PORT beneficiary in N1 stage.