BACKGROUND: Empagliflozin has been proposed as a treatment for COVID-19 on the basis of its anti-inflammatory, metabolic, and haemodynamic effects. The RECOVERY trial aimed to assess its safety and efficacy in patients admitted to hospital with COVID-19. METHODS: In the randomised, controlled, open-label RECOVERY trial, several possible treatments are compared with usual care in patients hospitalised with COVID-19. In this analysis, we assess eligible and consenting adults who were randomly allocated in a 1:1 ratio to either usual standard of care alone or usual standard of care plus oral empagliflozin 10 mg once daily for 28 days or until discharge (whichever came first) using web-based simple (unstratified) randomisation with allocation concealment. The primary outcome was 28-day mortality; secondary outcomes were duration of hospitalisation and (among participants not on invasive mechanical ventilation at baseline) the composite of invasive mechanical ventilation or death. On March 3, 2023 the independent data monitoring committee recommended that the investigators review the data and recruitment was consequently stopped on March 7, 2023. The ongoing RECOVERY trial is registered with ISRCTN (50189673) and ClinicalTrials.gov (NCT04381936). FINDINGS: Between July 28, 2021 and March 6, 2023, 4271 patients were randomly allocated to receive either empagliflozin (2113 patients) or usual care alone (2158 patients). Primary and secondary outcome data were known for greater than 99% of randomly assigned patients. Overall, 289 (14%) of 2113 patients allocated to empagliflozin and 307 (14%) of 2158 patients allocated to usual care died within 28 days (rate ratio 0·96 [95% CI 0·82-1·13]; p=0·64). There was no evidence of significant differences in duration of hospitalisation (median 8 days for both groups) or the proportion of patients discharged from hospital alive within 28 days (1678 [79%] in the empagliflozin group vs 1677 [78%] in the usual care group; rate ratio 1·03 [95% CI 0·96-1·10]; p=0·44). Among those not on invasive mechanical ventilation at baseline, there was no evidence of a significant difference in the proportion meeting the composite endpoint of invasive mechanical ventilation or death (338 [16%] of 2084 vs 371 [17%] of 2143; risk ratio 0·95 [95% CI 0·84-1·08]; p=0·44). Two serious adverse events believed to be related to empagliflozin were reported: both were ketosis without acidosis. INTERPRETATION: In adults hospitalised with COVID-19, empagliflozin was not associated with reductions in 28-day mortality, duration of hospital stay, or risk of progressing to invasive mechanical ventilation or death so is not indicated for the treatment of such patients unless there is an established indication due to a different condition such as diabetes. FUNDING: UK Research and Innovation (Medical Research Council) and National Institute of Health Research (MC_PC_19056), and Wellcome Trust (222406/Z/20/Z). TRANSLATIONS: For the Nepali, Hindi, Indonesian (Bahasa) and Vietnamese translations of the abstract see Supplementary Materials section.