Summary: Background: SARS-CoV-2 infections still have a significant impact on the global population. The existing vaccinations have contributed to reducing the severe disease courses, decreasing hospitalisations, and lowering the mortality rate. However, due to the variability of COVID-19 symptoms, the emergence of new variants and the uneven global distribution of vaccines there is still a great need for new therapy options. One promising approach is provided by host-directed therapies. We assessed here the efficacy and safety of MP1032, a host-directed anti-viral/anti-inflammatory drug in hospitalised patients with moderate to severe COVID-19. Methods: In a randomised, double-blind, placebo-controlled, Phase IIa study, patients were randomised 2:1 to receive either 300 mg MP032 bid + Standard-of-Care (SoC) or placebo bid + SoC for 28 days. Eligible patients were ≥18 years old, tested positive for SARS-CoV-2, and had moderate to severe COVID-19 symptoms. The study spanned 20 sites in six countries (Bulgaria, France, Hungary, Italy, Romania, Spain), assessing disease progression according the NIAID scale as the primary outcome on day 14. Secondary objectives included disease progression (day 28), disease resolution (days 14 and 28), mortality rate, COVID-19 related parameters and safety. Exposure-response analyses were performed, linking MP1032 to COVID-19 biomarkers (eGFR, D-dimer). Findings: 132 patients were enrolled to receive MP1032 + SoC (n = 87) or placebo + SoC (n = 45). The patients were all white or Caucasian with a mean (median) age of 60.5 (63) years. Overall, only 10 patients were vaccinated, 5 in each group. No significant risk difference of disease progression could be detected between groups on both day 14 (9.8% MP1032 vs. 11.6% placebo) and day 28 with MH common risk differences of −0.276% (95% CI, −11.634 to 11.081; p = 0.962) and 1.722% (95% CI, −4.576 to 8.019; p = 0.592), respectively.The treatment with MP1032 + SoC was safe and well-tolerated. Overall, 182 TEAEs including 10 SAEs were reported in 53.5% (46/86) of patients of the verum group and in 57.8% (26/45) of patients of the placebo group; the SAEs occurred in 5.8% (5/86) and 6.7% (3/45) of verum and placebo patients, respectively. None of the SAEs was considered as related. Interpretation: Despite the study's limitation in size and the variation in concurrent SoCs, these findings warrant further investigation of MP1032 as a host-directed anti-viral drug candidate. Funding: The study was funded by the COVID-19 Horizon Europe work programme and MetrioPharm AG.