Background During the COVID-19 first wave in France, the capacity of intensive care unit (ICU) beds almost doubled, mainly because of the opening of temporary ICUs with staff and equipment from anaesthesia. Objectives We aim to investigate if the initial management in temporary ICU is associated with a change in ICU mortality and short-term prognosis. Design Retrospective single-centre cohort study. Setting Surgical ICU of the Bichat Claude Bernard University Hospital during the COVID-19 “first wave” (from 18 March to 10 April 2020). Patients All consecutive patients older than 18 years of age with laboratory-confirmed SARS-CoV-2 infection and/or typical radiological patterns were included during their first stay in the ICU for COVID-19. Intervention Patients were admitted to a temporary ICU if no room was available in the classical ICU and if they needed invasive mechanical ventilation but no renal replacement therapy or Extracorporeal Membrane Oxygenation (ECMO) in the short term. The temporary ICUs were managed by mixed teams (from the ICU and anaesthesiology departments) following a common protocol and staff meetings. Main outcome measure ICU mortality Results Among the 59 patients admitted, 37 (62.7%) patients had initial management in the temporary ICU. They had the same characteristics on admission and the same medical management as patients admitted to the classical ICU. ICU mortality was similar in the 2 groups (32.4% in temporary ICUs versus 40.9% in classical ICUs; p=0.58). SAPS-II and ECMO use were associated with mortality in multivariate analysis but not admission to the temporary ICU. Conclusion In an overload context of the ICU of a geographical area, our temporary ICU model allowed access to intensive care for all patients requiring it without endangering them.