Introduction and aim. The aim of this study was to examine the usability of blood urea nitrogen to serum albumin ratio (BAR) as a prognostic predictor of in-hospital mortality in patients with gastrointestinal (GI) bleeding. Material and methods. The electronic medical records of the patients who applied to the emergency department due to upper GI bleeding during the study period were reviewed. The receiver operating characteristic (ROC) curve and the area under the curve (AUC) were used to evaluate each discriminant cut-off value to estimate mortality. Results. The study included 225 patients. The median (IQR) age of the patients was 75.0 (68.0–84.0) and 94 (41.8%) were female. AUC was determined as 0.784±0.055 (95% CI, 0.677–0.892) for BAR (p<0.001) in terms of in-hospital mortality. The cut-off value of BAR for this outcome was calculated as 16.26. In this cut-off value, sensitivity was 71.43%, specificity 82.84%, positive predictive value (PPV) 30.00% and negative predictive value (NPV) 96.57%. Conclusion. BAR is a useful tool that can be used to predict the in-hospital mortality of patients with GI bleeding. Patients with GI bleeding with a BAR above 16.26 will require more aggressive and timely intervention.