Comparison of measured overtopping data with computations with Swash. The SWASH output discharges are smaller than the 5% lower limit of Eurotop and Neural Network predictions. Nevertheless, they were close values and they have the same order of magnitude, which is very important when comparing overtopping values. No refection analysis was done. The smaller discharges are believed to be explained by the fact that SWASH does not describe splash, and therefore, only overtopping due to waves running up the slope, but not due to “spray” is taken into account. In addition to this, it is thought that SWASH does not model accurately enough wave breaking over abrupt changes in bottom geometry or steep slopes resulting in underestimation of overtopping. On the other hand, SWASH is not able to predict wave overtopping at rubble mound breakwaters. A breakwater is not well modelled in SWASH, because porous structures are dealt as a numerical dissipation box and not as a physical obstacle for incoming waves. As a result, waves are damped but not diverted upwards, so there is no overtopping. The reasons why this happens and some recommendations are given in this thesis. Besides, some ideas about how to introduce a multi-layer structure are explained.