This study aims to determine cutoff values for shock index (SI) to predict the need for transfusion and composite adverse outcomes in postpartum hemorrhage (PPH) cases. One hundred thirty PPH cases (study group) that necessitated blood transfusion were retrospectively compared to a frequency-matched control group (n = 130). Receiver operating characteristic (ROC) curves and decision tree [Classification & Regression Tree (C&RT) and Chi-square Automatic Interaction Detector (CHAID)] were used to identify cutoff values for SI. Cutoff values for postdelivery, peak and delta SI values for the prediction of PPH that required transfusion were 0.9125 (0.815 sensitivity, 0.923 specificity), 0.9145 (0.892 sensitivity, 0.823 specificity) and 0.195 (0.823 sensitivity, 0.885 specificity), while cutoff values for the same SI values in the prediction of composite adverse outcome were 1.315 (0.645 sensitivity, 0.616 specificity), 1.183 (0.613 sensitivity, 0.737 specificity) and 0.487 (0.710 sensitivity, 0.758 specificity). Delta SI was superior to postdelivery and peak SI in the prediction of PPH that required transfusion. Peak SI was superior to postdelivery and delta SI in the prediction of composite adverse outcome. In conclusion, increased postdelivery, peak, and delta SI values were related to adverse outcomes for PPH. SI seems to be a practical and effective method for the objective assessment of postpartum hemorrhage.