In the maritime industry, cavitation erosion prediction becomes more and more critical, as the requirements for more efficient propellers increase. Model testing is yet the most typical way a propeller designer can, nowadays, get an estimation of the erosion risk on the propeller blades. However, cavitation erosion prediction using computational fluid dynamics (CFD) can possibly provide more information than a model test. In the present work, we review erosion risk models that can be used in conjunction with a multiphase unsteady Reynolds‐averaged Navier–Stokes (URANS) solver. Three different approaches have been evaluated, and we conclude that the energy balance approach, where it is assumed that the potential energy contained in a vapor structure is proportional to the volume of the structure, and the pressure difference between the surrounding pressure and the pressure within the structure, provides the best framework for erosion risk assessment. Based on this framework, the model used in this study is tested on the Delft Twist 11 hydrofoil, using a URANS method, and is validated against experimental observations. The predicted impact distribution agrees well with the damage pattern obtained from paint test. The model shows great potential for future use. Nevertheless, it should further be validated against full scale data, followed by an extended investigation on the effect of the driving pressure that leads to the collapse.