We conducted a systematic review aiming to summarize the data on the current hemorrhage prediction models and evaluate their potential for generalized application in the neonatal population. The electronic databases PubMed and Scopus were searched, up to September 20, 2023, for studies that focused on development and/or validation of a prediction model for bleeding risk in neonates, and described the process of model building. Nineteen studies fulfilled the inclusion criteria for the present review. Eighteen bleeding risk prediction models in the neonatal population were identified, four of which were internally validated, one temporally and one externally validated. The existing prediction models for neonatal hemorrhage are mostly based on clinical variables and do not take into account the clinical course and hemostatic profile of the neonates. Most studies aimed at predicting the risk of intraventricular hemorrhage (IVH) reflecting the fact that IVH is the most frequent and serious bleeding complication in preterm neonates. A justification for the study sample size for developing the prediction model was given only by one study. Prediction and stratification of risk of hemorrhage in neonates is yet to be optimized. To this end, qualitative standards for model development need to be further improved. The assessment of the risk of bleeding incorporating platelet count, coagulation parameters, and a set of relevant clinical variables is crucial. Large, rigorous, collaborative cohort studies are warranted to develop a robust prediction model to inform the need for transfusion, which is a fundamental step towards personalized transfusion therapy in neonates.