Estimation of wave overtopping over the crest of coastal structures is crucial to design effective and cost efficient countermeasures against storms. Semi-empirical formulas are often used for wave overtopping assessment, but they are not always applicable for complex structures which exist in reality (e.g. a storm wall on a dike in a shallow or very shallow foreshore). Detailed numerical models such as Eulerian and Lagrangian RANS models have potential to simulate overtopping of complex coastal structures with good accuracy. However such models require significant computational resources. The use of such models is often not feasible for the design of coastal structures, which often requires multiple iterations and model runs over a reasonably long period of time (e.g. wave trains with 1000 individual waves). In this paper we investigated the applicability of the simplified depth integrated wave transformation model SWASH for wave overtopping estimation of impermeable coastal structures in shallow foreshores. The validation results demonstrate the capability of SWASH to predict mean wave overtopping discharge with good accuracy compared to results from four different overtopping experimental campaigns (comprising 124 individual cases). The overall performance of SWASH to estimate mean wave overtopping discharge is as accurate as those obtained by semi-empirical equations in literature. However, in order to obtain accurate mean wave overtopping discharge with the SWASH model, the incident wave properties at the toe of the dike need to be accurately reproduced. For cases where this is not possible, a correction method is proposed in this paper. Detailed validation of the instantaneous wave overtopping also shows a good agreement with physical model data. In one example, a single, intensive overtopping event was not well resolved by the SWASH model and the instantaneous wave overtopping was under-predicted. However, this did not contribute significantly to the mean wave overtopping discharge. An additional advantage of the SWASH model is that specific coastal structure geometries can be modelled in SWASH if they are not covered by semi-empirical equations. Even in a case with rapidly varied flow (e.g. vertical wall on a dike) the model shows sufficient robustness. In this paper the details on the SWASH model configuration and post processing methods are outlined to enable the reader to reproduce reliable wave overtopping estimation over impermeable coastal structures in shallow foreshores.@en