Objective <p>Acute heart failure (AHF) is a common complication in critically ill patients. The purpose of this study was to investigate the relationship between blood urea nitrogen to albumin ratio (BAR) and poor prognosis of AHF in the Intensive Care Unit (ICU). </p> Methods <p>AHF patients were selected from the MIMIC-IV 2.2 database and data were extracted. Boruta algorithm was used for feature selection. Cox regression analysis, restricted cubic spline (RCS) analysis and Kaplan-Meier survival curve were used to evaluate the association between BAR and mortality. Receiver Operating Characteristic (ROC) was used to evaluate the diagnostic accuracy and subgroup analysis was performed to evaluate the universality of the results. </p> Results <p>A total of 671 patients were included in the database, and Boruta analysis identified BAR as an important clinical feature. BAR, whether as a continuous variable or grouped, suggests that BAR is a predictor of poor prognosis of AHF, and a higher level indicates a poor prognosis. The RCS model reveals the nonlinear relationship between BAR and mortality. The Area Under Curve (AUC) and 95% CI of BAR in predicting 28-day mortality and 90-day mortality were 0.645 (0.588–0.699) and 0.643 (0.593–0.692), respectively. Subgroup analysis showed that age, gender, Continuous Renal Replacement Therapy (CRRT) and scoring system were the factors affecting the predictive value.</p> Conclusion <p>BAR may have potential value in predicting the mortality of patients with AHF in ICU, supporting the early use of risk assessment. However, further studies are needed to confirm this result.</p>