In recent years OpenFOAM studies have focused on pure wave modelling and wave-structure interaction. Considering that there is a lack of transparency concerning effects of different motion modelling settings in OpenFOAM, this paper presents the influence of different numerical setups on the accurate modelling of wave-induced motions. This is achieved by applying the interFoam solver included in OpenFOAM-v2206 with the waves2Foam toolbox. The optimal numerical setup is studied with a two-dimensional box-like ship idealisation heaving in varying wave lengths while the wave steepness remains constant. Different numerical setups are considered for accurate wave modelling, the modelling of wave excitation forces for the case of a static structure, and heave motion modelling. Finally, the optimal setup that is found is applied for a 3D ship case in head waves with heave and pitch coupling. As far as practically possible, the results are validated against experimental data. It is shown that strict requirements for mesh density and time step for accurate wave modelling result in accurate excitation force and motion results. Reflections from the relaxation zones and different mesh density layers cause inaccuracies. It is concluded that simple numerical cases are suitable for studying optimal numerical setups for more complex simulation cases. •Ten cells per wave height and an aspect ratio of four are accurate for wave modelling.•k−ω SST turbulence modelling is problematic in model-scale waves.•An upwind divergence scheme stabilises force and motion modelling results.•The dense resolution needed for wave modelling is accurate for wave structure studies.•Optimal numerical setups for complex cases can be studied with simple 2D cases.