Background : The prognostic nutrition index (PNI) has been shown to have prognostic value for several common cancers. The study aim was to explore the clinical application value of the PNI for prognosis of patients with oesophageal squamous cell carcinoma (ESCC) treated with radical chemoradiotherapy (CRT) or radiotherapy (RT). Methods : Overall, 193 patients with ESCC who received radiotherapy with or without chemotherapy at Sichuan Cancer Hospital from March 20, 2012 to December 25, 2017 were retrospectively analysed. Based on serum measurements before treatment, the PNI at ESCC recurrence was calculated as albumin (g/L) + 5 × total lymphocyte count. The Kaplan–Meier method and Cox proportional regression model were used to analyse the relationship between PNI and overall survival (OS). Results : The average pretreatment PNI of 193 ESCC patients was 49.01 ± 4.68. The optimal cutoff value of PNI was 47.975, and the patients were divided into a low-PNI group (<47.975) and a high-PNI group (≥47.975). PNI was related to tumour length, T-stage and synchronous chemotherapy in ESCC patients (P < 0.05). The median OS for the entire group was 22.37 months. The median OS of patients in the high-PNI group and low-PNI group were 32.63 months and 15.4 months, respectively, the 3-year survival rates were 47.5% and 32.2% and the 5-year survival rates were 37.7% and 16.8%, respectively, (all P = 0.001). Univariate analysis showed that PNI, tumour length, T-stage and synchronous chemotherapy were related to the prognosis of ESCC patients (P < 0.05). Multivariate analysis showed that tumour length (P = 0.019), synchronous chemotherapy (P = 0.009) and PNI (P = 0.003) were independent prognostic factors affecting the prognosis of patients in ESCC treated with RT or CRT. Conclusions: The calculation of PNI value is simple, reliable and repeatable and can improve the accuracy of a patient’s prognosis. Confirmation of these results by a large-sample prospective study is desirable.