A crown wall is a gravity based L- shaped concrete structure on top of a rubble mound breakwater. Functions of a crown wall are: reduction of wave overtopping, enable transport on the breakwater and placement of pipelines on top of the flat surface. Marine contractor Van Oord was awarded to extend a breakwater with a crown wall on top, located in Romania. The design, already made by a local consultant, was reviewed by Van Oord, which led to the conclusion that the crown wall would be blown of the breakwater for that particular design in combination with the design wave conditions. Subsequently, physical scaled model tests were executed by Artelia in Grenoble, but, against all expectations the crown wall appeared to remain stable. Based on the contradiction between stability calculations and the outcome of the scaled model tests, the hypothesis arose that currently used wave load calculation methods lead to designs that seem to be conservative especially when freeboard, the vertical distance between the still water level and the base of the crown wall, increases. An experimental research is carried out in which the effect of freeboard is examined by varying water levels and wave conditions while geometrical properties of the structure remain constant. Tests are divided in three main subjects: pressure measurements, uplift stability and overall stability. Based on pressure measurements it is concluded that the mostly used design method PEDERSEN [1996] and its extended version of NØRGAARD et al. [2013], assume an upward pressure distribution which is too conservative in shape and distance over which pressure is exerted against the base of the crown wall. These methods assume a linear pressure distribution whereas this seems to be conservative based on test results since S-shaped and parabolic upward pressure distributions are found. Furthermore, it is assumed that upward pressure acts over the full length of the base whatever wave conditions and freeboard are. However, a relation is found which indicates that the effective length of the upward pressure actually depends on wave height and freeboard. Comparing the predicted vertical loads based on the methods of PEDERSEN [1996] and NØRGAARD et al. [2013] to the found vertical loads derived from uplift stability in this research, lead to substantial overpredictions when freeboard increases. Generated data with respect to critical weights confirms that the conventional methods of PEDERSEN [1996] and NØRGAARD et al. [2013] are too conservative in predicting overall stability with respect to sliding, especially when freeboard increases. However, by implementing the found relations for the effective length, especially NØRGAARD et al. [2013] become more reliable with respect to the found critical weights. Design guidelines are presented. It is advised to use these guidelines for prior design purposes in combination with the adapted method of NØRGAARD et al. [2013] in which a reduction coefficient for vertical loads must be taken into account. Since the range of application for the critical weight predictions is still small it is recommended to extend this range by varying more parameters in further research, which should make these guidelines generally better applicable.