Purpose:Endostatin is a promising biomarker for predicting acute kidney injury (AKI) and mortality in the intensive care unit (ICU). We investigated plasma endostatin upon ICU admission as a predictor of AKI, renal replacement therapy (RRT), and 30-day mortality.Methods:A retrospective multicenter study was performed with admissions (ICU length of stay ≥24 hours) to four ICUs. KDIGO criteria defined AKI. Endostatin on ICU admission was compared to creatinine, cystatin C, and the Simplified Acute Physiology Score 3 (SAPS-3). Admissions with sepsis and creatinine <100 μmol/L on ICU admission underwent subgroup analyses. Regression models and the area under the receiver operating characteristic curve (AUC) were assessed.Results:In total, 4449 admissions were included (43% sepsis and 61% AKI). Endostatin was associated with AKI (odds ratio [OR] 1.6, 95% confidence interval [CI] 1.4-1.7), future AKI (OR 1.5, 95% CI 1.4-1.7), future AKI stage 3 (OR 1.4, 95% CI 1.2-1.6), and RRT (OR 1.2, 95% CI 1.1-1.4) independently of creatinine and cystatin C, with similar results in sepsis. Endostatin was also associated with time to AKI (hazard ratio 1.2, 95% CI 1.1-1.2). For admissions with creatinine <100 μmol/L, endostatin (AUC 0.62, 95% CI 0.59-0.65) outperformed creatinine (AUC 0.51, 95% CI 0.49-0.54) and cystatin C (AUC 0.53, 95% CI 0.50-0.56) in predicting future AKI (p<0.001). Endostatin was not associated with 30-day mortality after adjusting for SAPS-3.Conclusion:Endostatin is an early and potentially clinically useful biomarker for predicting AKI and RRT needs at ICU admission, especially in patients with low to mildly elevated creatinine.