Severe alcohol-associated hepatitis (SAH) is associated with high short-term mortality, and failure of response to corticosteroids is associated with a mortality of ~70%-80% within 6 months. Granulocyte colony-stimulating factor (G-CSF) has been studied in steroid non-responders; however, the data are limited. This is a multicentre retrospective cohort study. The study period was from January 2016 to November 2023. SAH was defined as alcohol-associated hepatitis (ICD-10-CM codes) with serum bilirubin ≥ 5.0 mg/dL and INR ≥ 1.5. Other aetiologies of acute hepatitis and biliary obstruction were excluded. The primary outcome was 90-day median overall survival in SAH patients treated with G-CSF compared with standard medical therapy (SMT) or corticosteroids. Propensity score (1:1) matching was performed to control confounding variables. Among 20 132 patients with SAH, 10800 (53.65%) were treated with corticosteroids and 224 (1.11%) G-CSF. The G-CSF group was younger (45.5 vs. 48.4) White (79.91% vs. 72.40%); however, there was no age or gender difference between G-CSF and corticosteroid groups. Whites and patients with more comorbidities received G-CSF more frequently than SMT or corticosteroids. After propensity score matching, 90-day overall survival was better in patients who received G-CSF (88.31% vs. 62.36%, p < 0.01) compared with SMT or corticosteroids (88.31% vs. 74.39%, p < 0.01). Patients on G-CSF had better 6-month transplant-free survival compared with SMT (83.53% vs. 55.36%, p < 0.001) or corticosteroids (82.89% vs. 60.21%, p < 0.001). Gastrointestinal bleeding was less common in G-CSF group compared with corticosteroids (5.02% vs. 10.50%, p < 0.001). A small minority of patients with severe alcohol-associated hepatitis receive G-CSF. G-CSF improves 90-day overall survival in patients with severe alcohol-associated hepatitis and is non-inferior to corticosteroids.