BACKGROUND AND AIM: Short-term mortality in alcohol-related hepatitis (AH) is high and no current therapy results in durable benefit. A role for IL-1ß has been demonstrated in the pathogenesis of alcohol-induced steatohepatitis. This study explored the safety and efficacy of canakinumab (CAN), a monoclonal antibody targeting IL-1ß, in the treatment of patients with AH. METHODS: Participants with biopsy-confirmed AH and discriminant function ≥32 but MELD ≤27 were randomly allocated 1:1 to receive either CAN 3mg/kg or placebo (PBO). Liver biopsies were taken before, and 28 days after treatment. The primary endpoint was the overall histological improvement in inflammation analysed by modified Intention-To-Treat (mITT). RESULTS: Fifty-seven participants were randomised: 29 to CAN and 28 to PBO. Two participants had histology that did not corroborate the clinical diagnosis. Of the remaining 55 participants, paired histology data was evaluable from 48 participants. In CAN-treated participants, 14/24 (58%) demonstrated histological improvement compared to 10/24 (42%) in the PBO group (p=0.25). There was no improvement in prognostic scores of liver function. Four of the 55 participants (7%) died within 90 days; 2 in each group. The number of serious adverse events was similar between CAN vs PBO. In post-hoc exploratory analyses after adjustment for baseline prognostic factors, CAN therapy was associated with overall histological improvement (p=0.04). CONCLUSION: CAN therapy in severe AH participants with MELD≤27 did not alter biochemical or clinical outcomes compared to PBO. Non-significant histological improvements did not translate into clinical benefit.