Rip currents are among the most dangerous coastal hazards for the bathing public, and contribute to the highest portion of beach rescues all over the world. In order to help life guards in planning and preparing rescue resources so that casualties can be minimized, information about where and when rip currents may occur is needed. This can be provided by a predictive tool which combines meteorological forecasts, hydrodynamic models and remote-sensed observations. In this thesis, a methodology which can provide rip current forecasts for swimmer safety is developed and tested for Egmond aan Zee beach in the Netherlands. The approach uses the numerical model system CoSMoS, combined with daily estimates of nearshore-scale bathymetry obtained from a system called cBathy, which infers depths by estimating wave celerities from video imaging. Furthermore, in order to gain more knowledge on occurrences of rips at Egmond beach, a numerical study on the kinematics of rip currents and the safety implications for swimmers is presented as well. Coupling the video bathymetry estimates with CoSMoS in forecast mode shows that dangerous rips were correctly predicted. This thesis demonstrates the potential application of the proposed system for providing rip current forecasts at Egmond aan Zee.