Abstract Low Vitamin D (VD) levels in COVID-19 patients have been related to increased disease severity and worse outcomes. However, most of these trials designs were not-controlled and retrospective, including patients with demographic differences and biases potentially influencing the reported associations between lower VD levels and severe COVID-19. Aim of this study was to prospectively evaluate VD levels influence on disease severity in a COVID-19 patients cohort matched for age, sex and comorbidities with control subjects. Patients admitted to San Raffaele University Hospital for COVID-19 from March to June 2021 were consecutively enrolled in this study, which was approved by the local IRB, after giving their informed consent. Severe (i. e. those needing high flow oxygen therapy) and non-severe COVID-19 patients matched for age, sex and comorbidities were recruited at admission in Emergency Department (ED). Control subjects from the outpatient Endocrinology Unit of the same Hospital were enrolled in the same period. We excluded patients with comorbidities and therapies influencing VD metabolism. 25OH-VD levels were evaluated at admission in ED and VD deficiency was defined by serum 25OH-VD below 20 ng/mL. A total of 73 COVID-19 patients and 30 control subjects were included in the study. No differences regarding age, sex and comorbidities were found between patients and control groups, but 25OH-VD levels were statistically lower in COVID-19 patients (13.3 vs 23.6 ng/mL, p<0. 001) with a higher prevalence of VD deficiency (75% vs 43%, p=0. 002). After the ED admission, during the hospitalization, a severe disease occurred in a total of 46 COVID-19 patients, and 27 were affected by a non-severe one. No differences regarding age, sex and comorbidities were found between severe vs non-severe groups, but 25OH-VD was significantly lower in the severe one (11.7 vs 16.7 ng/mL, p=0. 007) with a higher prevalence of VD deficiency (85% vs 59%, p=0. 015). We observed lower 25OH-VD levels at admission in patients with non-invasive mechanical ventilation requirement (n.18) and in those admitted in intensive care unit (n.8) during hospitalization, compared to the other patients (p=0. 003 and p=0. 01), although no differences regarding age and comorbidities were found. Moreover, in patients with VD deficiency we found higher levels of inflammatory markers C-reactive protein and ferritin (69 vs 42 mg/L, p=0. 036; 723 vs 414 ng/mL, p=0. 028), and lower SaO2/FiO2 and PaO2/FiO2 ratios (433 vs 447, p=0. 005; 261 vs 311; p<0. 001) at hospital admission. In conclusion, our prospective data confirm that low VD levels are widely found in hospitalized COVID-19 patients compared to control subjects and predict increased disease severity independently from age, sex and comorbidities of patients affected. We suggest that tackling VD deficiency may be an effective preventive measure to prevent severe COVID-19. Presentation: Saturday, June 11, 2022 1:00 p.m. - 3:00 p.m.