The study developed and validated a prediction model for in-hospital mortality in intensive care unit patients with cirrhosis and sepsis. The data come from the MIMIC-IV (n=2052 patients), eICU-CRD (n=657) and SZPH-JIS (n=131) databases. The training cohort (2008–2016; n=1328) had a 24.0% mortality, the temporal validation (2017–2022; n=724) 35.9%, eICU-CRD 25.9% and SZPH-JIS 38.2%. The model includes 13 predictors: age, respiratory rate, body temperature, oxygen saturation, heart rate, total bilirubin, lactate, creatinine, white blood cell count, INR, use of vasopressors, urine output, and Glasgow Coma Scale score. It achieved an AUROC of 0.822 (95% CI: 0.797–0.847) in training, 0.810 (95% CI: 0.777–0.843) in temporal validation, 0.777 (95% CI: 0.734–0.821) in eICU-CRD, and 0.763 (95% CI: 0.734–0.821) in eICU-CRD. 0.680–0.846) in SZPH-JIS. The model has demonstrated better discriminatory performance than existing prognostic scores and is available as a web-based risk stratification calculator.