The clinical picture of SARS-CoV-2 infection is variable, ranging from moderate to severe and critical forms. Various studies have shown that disturbances in certain biochemical parameters are correlated with the severity of the disease. They can be used to predict the risk of complications, allowing early and appropriate care of these patients. The aim of our study is to evaluate the correlation between admission ferritin levels and mortality in COVID-19 patients. We used data from the records of patients with positive RT-PCR admitted to the Mohamed V Military Training Hospital in Rabat between april 2020 and december 2021. Demographic and biologic data were exctracted from DxLab LIS of Dedalus. 1.310 patients are included; the average age is 56 years and 72% of patients are male. The all-cause mortality of seriously ill hospitalized patients in the intensive care units are 23%. 73% of patients who died were over 60 years. The mean ferritinemia at admission was 2040 ng/ml for patients who died versus 521 ng/ml for the survivors (p <0.01). The AUC of the ROC curve for ferritin levels predicting all-cause mortality is 0.745. The threshold of 800 ng/ml allows a sensitivity of 72% and a specificity of 30%. Logistic regression analysis showed that ferritin levels above 800 ng/ml and age above 60 years are risk factors for mortality (p<0.01) independently of each other. At hospitalisation of COVID-19 patients, high ferritin levels would be correlated with a bad prognosis justifying an early and adapted care of these patients.