Patients with locoregionally advanced nasopharyngeal carcinoma have a high risk of disease relapse, despite a high proportion of patients attaining complete clinical remission after receiving standard-of-care treatment (ie, definitive concurrent chemoradiotherapy with or without induction chemotherapy). Additional adjuvant therapies are needed to further reduce the risk of recurrence and death. However, the benefit of adjuvant chemotherapy for nasopharyngeal carcinoma remains controversial, highlighting the need for more effective adjuvant treatment options. This multicentre, open-label, parallel-group, randomised, controlled, phase 3 trial was done at 14 hospitals in China. Patients (aged 18-65 years) with histologically confirmed, high-risk locoregionally advanced nasopharyngeal carcinoma (stage III-IVA, excluding T3-4N0 and T3N1 disease), no locoregional disease or distant metastasis after definitive chemoradiotherapy, an Eastern Cooperative Oncology Group performance status of 0 or 1, sufficient haematological, renal, and hepatic function, and who had received their final radiotherapy dose 12-16 weeks before randomisation, were randomly assigned (1:1) to receive either oral metronomic capecitabine (650 mg/m2 body surface area twice daily for 1 year; metronomic capecitabine group) or observation (standard therapy group). Randomisation was done with a computer-generated sequence (block size of four), stratified by trial centre and receipt of induction chemotherapy (yes or no). The primary endpoint was failure-free survival, defined as the time from randomisation to disease recurrence (distant metastasis or locoregional recurrence) or death due to any cause, in the intention-to-treat population. Safety was assessed in all patients who received at least one dose of capecitabine or who had commenced observation. This trial is registered with ClinicalTrials.gov, NCT02958111. Between Jan 25, 2017, and Oct 25, 2018, 675 patients were screened, of whom 406 were enrolled and randomly assigned to the metronomic capecitabine group (n=204) or to the standard therapy group (n=202). After a median follow-up of 38 months (IQR 33-42), there were 29 (14%) events of recurrence or death in the metronomic capecitabine group and 53 (26%) events of recurrence or death in the standard therapy group. Failure-free survival at 3 years was significantly higher in the metronomic capecitabine group (85·3% [95% CI 80·4-90·6]) than in the standard therapy group (75·7% [69·9-81·9]), with a stratified hazard ratio of 0·50 (95% CI 0·32-0·79; p=0·0023). Grade 3 adverse events were reported in 35 (17%) of 201 patients in the metronomic capecitabine group and in 11 (6%) of 200 patients in the standard therapy group; hand-foot syndrome was the most common adverse event related to capecitabine (18 [9%] patients had grade 3 hand-foot syndrome). One (<1%) patient in the metronomic capecitabine group had grade 4 neutropenia. No treatment-related deaths were reported in either group. The addition of metronomic adjuvant capecitabine to chemoradiotherapy significantly improved failure-free survival in patients with high-risk locoregionally advanced nasopharyngeal carcinoma, with a manageable safety profile. These results support a potential role for metronomic chemotherapy as an adjuvant therapy in the treatment of nasopharyngeal carcinoma. The National Natural Science Foundation of China, the Key-Area Research and Development Program of Guangdong Province, the Natural Science Foundation of Guangdong Province, the Innovation Team Development Plan of the Ministry of Education, and the Overseas Expertise Introduction Project for Discipline Innovation. For the Chinese translation of the abstract see Supplementary Materials section.