We aimed to investigate patterns of corticosteroid use and their relationship with remdesivir use and clinical outcomes in a large real-life cohort of COVID-19 patients treated in a tertiary-level institution. We retrospectively analyzed a total of 1558 severe and critical COVID-19 patients, including 779 patients treated with remdesivir and 779 matched control patients. A total of 167 (10.7%) patients received none, 710 (45.6%) low, 539 (34.6%) high, and 142 (9.1%) very high corticosteroid doses. Patients treated with remdesivir had significantly longer exposure to corticosteroids, received higher average and maximal daily doses, and cumulative corticosteroid doses. In the multivariate analysis remdesivir use, lower cumulative comorbidity burden, higher severity of COVID-19 symptoms, and mechanical ventilation were recognized as mutually independent predictors of the use of higher corticosteroid doses. Higher corticosteroid doses were associated with significantly increased mortality.Among non-remdesivir treated patients, there was a U-shaped relationship between maximal daily corticosteroid dose and mortality. Among remdesivir treated patients gradual increase in mortality with increasing corticosteroid doses was observed. Patterns of corticosteroid use differ regarding the use of remdesivir and may moderate its association with survival among severe and critical COVID-19 patients.