Background To test whether known prognosticators of COVID-19 maintained their stratification ability across age groups.Methods We performed a retrospective study. We included all patients (n = 2225), who presented to the Emergency Department of the Careggi University Hospital for COVID-19 in the period February 2020-May 2021, and were admitted to the hospital. The following parameters were analyzed as dichotomized: 1) SpO(2)/FiO(2) <= or > 214; 2) creatinine < or >= 1.1 mg/dL; 3) Lactic dehydrogenase (LDH) < or >= 250 U/mL; 4) C Reactive Protein (CRP) < or >= 60 mg/100 mL. We divided the study population in four subgroups, based on the quartiles of distribution of age (G1 18-57 years, G2 57-71 years, G3 72-81 years, G4 > 82). The primary end-point was in-hospital mortality.Results By the univariate analysis, the aforementioned dichotomized variables demonstrated a significant association with in-hospital mortality in all subgroups. We introduced them in a multivariate model: in G1 SpO2/FiO2 <= 214 (Relative Risk, RR 15.66; 95%CI 3.98-61,74), in G2 creatinine >= 1.1 mg/L (RR 2.87, 95%CI 1.30-6.32) and LDH >= 250 UI/L (RR 8.71, 95%CI 1,15-65,70), in G3 creatinine >= 1.1 mg/L (RR 1.98, 95%CI 1,17-3.36) and CRP >= 60 ng/L (RR 2.14, 95%CI 1.23-3.71), in G4 SpO(2)/FiO(2) <= 214 (RR 5.15, 95%CI 2.35-11.29), creatinine >= 1.1 mg/L (RR 1.75, 95%CI 1.09-2.80) and CRP >= 60 ng/L (RR 1.82, 95%CI 1.11-2.98) were independently associated with an increased in-hospital mortality.Conclusions A mild to moderate respiratory failure showed an independent association with an increased mortality rate only in youngest and oldest patients, while kidney disease maintained a prognostic role regardless of age.