Moored vessels are subject to wave forces and moments at different frequencies, which induce motions of the body and can be transferred to the mooring lines and fenders. Under extreme forcing and vessel motions, dangerous line breaking accidents can occur or the ship movements can be simply too large to continue the on/off-loading process, causing downtime of the port. An accurate modelling of the waves in the harbour and the response of moored ships is of prime importance to determine the safety and workability of the berths. In this project the combination of the wave model SWASH and the 3D diffraction model Harberth to compute waves forces acting on moored ships is investigated. The proposed approach using the SWASH wave model come as a possible alternative to the current practice of Royal HaskoningDHV, who experienced a sequence of numerical issues while simulating extreme incident waves with an operational Boussinesq-type model. The objectives of the study lead to the following research question: To which extent can the SWASH wave model and the Harberth model be combined in order to accurately compute wave forces acting on moored ships? Each computational tool considered in the proposed approach (the wave model SWASH, the coupling procedure, and the diffraction model Harberth) can contribute to the (in)accuracy of the computed wave forces acting on a moored ship. A series of systematic tests with SWASH indicates that the dispersion errors are generally negligible, while amplitude errors remain small provided that: i) the number of vertical layers is sufficient (especially in relatively deep waters); ii) the horizontal resolution of the computational grid is sufficient; iii) important terms of the momentum equation are modelled with higher-order numerical schemes (especially for non-linear waves in relatively deep waters). The possible wave amplitude errors impact directly the predicted wave forces acting on the ship. The coupling tool developed to combine the SWASH and Harberth models proved to be consistent for the simplified tested conditions. The first-order forces computed using SWASH and Harberth are generally well predicted, while larger deviations can occur for the second-order forces. These are partially related to the simplification of motions made by Harberth in the computation of second-order forces. The proposed approach using the SWASH and Harberth models, combined with the developed coupling tool, is validated using model test data of waves and forces acting on a restrained ship. In general the performance is satisfactory, even for very energetic conditions, implying that the investigated modelling approach is a potential alternative for future applications. Finally, a list of recommendations is provided, including possible improvements in the SWASH model and Harberth model, and additional tests to consolidate the proposed modelling approach.