Pediatric patients with multiple organ failures in the pediatric intensive care unit (PICU) are at a higher risk of mortality. Assessing the mortality risk when patients are admitted to PICU is important to allocate treatment and care properly. The aim of this study was to compare the performance of the PRISM IV score within the first 24 and 72 hours to predict mortality in the PICU. Demographic, clinical, and laboratory data were collected to compute the PRISM IV in the first 24 and 72 hours among critically ill pediatric patients in the PICU at H. Adam Malik General Hospital, Medan, Indonesia, from April 2021 to February 2022. The comparison of the PRISM IV scores and its components within the first 24 and 72 hours was analyzed using the Wilcoxon test, Student's independent t-test or McNemar test. The role of PRISM IV score in predicting mortality was assessed using the receiver operating characteristic (ROC) curve. Out of 35 pediatric patients, 17 (48.6%) of them died. Platelet count (=0.022), pCO (=0.026), HCO (=0.009), total CO (=0.015), and base excess (=0.001) were statistically different between 24 and 72 hours groups. The area under curve (AUC) for the first 24 hours using PRISM IV scores was 47.4% with =0.792 (95%CI, 27.7%-67.1%). Meanwhile, the AUC of 72 hours group was 65.4%, =0.121 (95%CI, 47.1%-83.6%). This study suggested that PRISM IV scores in the first 24 and 72 hours may not be a reliable screening tool for predicting mortality. However, further studies are suggested to validate these findings.