A main challenge port engineers had to tackle in recent years is wave penetration inside a harbour, as it determines vessels’ safe sailing and mooring, possibly causes unwanted vessel movements, and unequivocally regulates the execution of port operations. A physical scale model can describe wave penetration in a complete way. However, the construction of a physical scale model is expensive and time consuming. For this reason, nowadays several numerical models are used to describe wave penetration in ports, affected by multiple processes such as diffraction, partial reflection, etc. In this study, the simulation of wave penetration with the non-hydrostatic model SWASH is examined. To validate the numerical model, output of an open benchmark dataset of Deltares (Deltares, 2016) is used, consisting of physical scale model tests of schematic port layouts. As wave penetration is a summation of physical processes, each process should be described accurately by SWASH. This thesis focuses on assessing how SWASH models wave penetration per wave process, first separately by means of simplified models and then combined in a model describing the full harbour layout resulting to the final wave field inside a port. The main topics of interest of the formulated research questions are the ability of SWASH to simulate wave propagation, wave celerity and the effect of two dominant wave processes: reflection and diffraction. As the amount of processes influencing wave penetration increases for higher layout complexity, the research was targeted at the simplest port layout considered in the benchmark dataset (Deltares, 2016). Moreover, only regular waves were taken into account, as in this case the differences between the measurements and the computational results are most easily identified. To better understand the influence of reflection the waves, two simplified one-dimensional SWASH models were designed. The first model simulated reflection in front of a gravel slope, located outside the harbour basin, and the second model reflection in front of the harbour basin end, consisting of a gravel slope and a concrete quay wall behind it. The results suggest that outside the basin the reflection off the gravel slope has a minor effect in comparison to the reflection off a vertical quay wall. Inside the harbour basin, wave reflection played a dominant role on the resulting wave field there. The SWASH models were proven to be robust as the wave height in the computational domain did not change considerably for an increase or decrease of the porosity of a gravel slope by 10%. It should be emphasised that the standing wave heights were altering fast within a short horizontal distance. Therefore, the precise wave height values were strongly influenced by the exact location of the output points examined. The importance of diffraction inside the harbour was demonstrated by a simplified two-dimensional model, in which reflection off the harbour end was not included. The information that could be obtained from the measurements about the wave height changes due to diffraction was limited. However, the initial trends due to diffraction were also identified in SWASH. From the comparison of the wave height in the SWASH model, influenced only by diffraction, to the respective measured value, it was confirmed that the total measured wave penetration inside the harbour was significantly influenced by diffraction. The comparison of the measurements to the results of the final SWASH model, which included the full version of the simplest physical model, showed that the overall wave field pattern is in agreement. The numerical model was able to reproduce the diffraction and reflection patterns observed in the measurements. At many output locations in SWASH the measured wave height values were simulated with high accuracy. On the downside, at other locations the measured and the modelled wave height deviated significantly. The large deviations can be explained by the fact that the standing wave patterns change within a short distance and thus the wave height can vary significantly at the area close to a specific output point. It may be possible that the measured wave height at a specific point can be identified in SWASH in the region close to the exact point coordinates. All in all, it was concluded that for non-breaking, relatively low waves, with wave height to water-depth ratio lower than 0.2, the accuracy of SWASH in modelling the wave processes of reflection and diffraction is sufficiently well for engineering purposes. For relatively high waves and/or breaking waves, numerical instabilities were detected. It is assumed that the numerical instabilities can be attributed to the relatively low number of grid cells per wave length. This study advances our understanding of the wave penetration simulation in SWASH. The approach followed allows investigating the ability of the model to simulate, separately and combined, two wave processes which predominantly contribute to wave penetration in harbour: reflection and diffraction. With further validation to guarantee the model stability, the strategy of this thesis can be a useful tool to understand the performance of SWASH in modeling wave penetration per wave process and in combination. The knowledge obtained enlightens the possible reasons leading to deviations between the measurements and the model outputs. This can be valuable assistance in the course of further improving the model accuracy.