Prediction of bleeding risk in patients receiving antithrombotic treatment has been repeatedly attempted and resulted in multiple risk models. Risk stratification can be used to personalize antithrombotic treatment to reduce bleeding complications. Although these risk models are validated and incorporated in the current guidelines, the feasibility and effectivity in daily practice remains questionable. Personalized antithrombotic treatment may be complex due to the number of risk factors that should be considered. Furthermore, most risk scores show moderate predictive value when they are validated externally or for a different treatment indication. However, personalized treatment seems key to minimalize bleeding events and is therefore indicated in all patients receiving antithrombotics. To increase the effectivity, it is important to use the appropriate model for each treatment indication. Online tools and the implementation of risk scores in the electronic medical record could facilitate the use of these scores.