Patients with advanced or metastatic oesophageal squamous cell carcinoma have poor prognosis and few treatment options after first-line therapy. We aimed to assess efficacy and safety of the anti-PD-1 antibody camrelizumab versus investigator's choice of chemotherapy in previously treated patients.ESCORT is a randomised, open-label, phase 3 study of patients aged 18 to 75 years with a histological or cytological diagnosis of advanced or metastatic oesophageal squamous cell carcinoma done at 43 hospitals in China. Eligible patients had an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1, and had progressed on, or were intolerant to, first-line standard therapy. Patients were randomly assigned (1:1) to camrelizumab (200 mg every 2 weeks) or chemotherapy with docetaxel (75 mg/m2 every 3 weeks) or irinotecan (180 mg/m2 every 2 weeks), all given intravenously. Central randomisation was done using the Randomization and Trial Supply Management system with block size randomly generated as four or six and stratified by disease and ECOG performance status. The primary endpoint was overall survival, assessed in randomised patients who had received at least one dose of treatment. Safety was assessed in all treated patients. The trial is registered with ClinicalTrials.gov, NCT03099382, and is closed to new participants.From May 10, 2017, to July 24, 2018, 457 (75%) of 607 screened patients were randomly assigned to treatment, of whom 228 received camrelizumab treatment and 220 received chemotherapy. As of data cutoff on May 6, 2019, with a median follow-up time of 8·3 months (IQR 4·1-12·8) in the camrelizumab group and 6·2 months (3·6-10·1) in the chemotherapy group, median overall survival was 8·3 months (95% CI 6·8-9·7) in the camrelizumab group and 6·2 months (5·7-6·9) in the chemotherapy group (hazard ratio 0·71 [95% CI 0·57-0·87]; two-sided p=0·0010). The most common treatment-related adverse events of grade 3 or worse were anaemia (camrelizumab vs chemotherapy: six [3%] vs 11 [5%]), abnormal hepatic function (four [2%] vs one [<1%]), and diarrhoea (three [1%] vs nine [4%]). Serious treatment-related adverse events occurred in 37 (16%) of 228 patients in the camrelizumab group, and in 32 (15%) of 220 patients in the chemotherapy group. Ten treatment-related deaths occurred, seven (3%) in the camrelizumab group (three deaths from unknown causes, one enterocolitis, one hepatic function abnormal, one pneumonitis, and one myocarditis) and three (1%) in the chemotherapy group (two deaths from unknown causes, and one gastrointestinal haemorrhage).Second-line camrelizumab significantly improved overall survival in patients with advanced or metastatic oesophageal squamous cell carcinoma compared with chemotherapy, with a manageable safety profile. It might represent a potential option of standard second-line treatment for patients with oesophageal squamous cell carcinoma in China.Jiangsu Hengrui Medicine.