© 2020 Elsevier Ltd Existing recommended practices in the literature do not provide clear and concise guidance for the selection of the most suitable numerical modelling strategy for investigating the boundary layer around a ship at full scale. For example, the International Towing Tank Conference procedure for calculating the nominal wake fields of full-scale ships does not clearly specify which turbulence modelling approach should be used to accurately represent the near-wall flow in the ship's aft region. This paper presents a numerical approach that can accurately represent the boundary layer of full-scale ships. Three turbulence modelling strategies, suitable for the simulation of ship flows, have been assessed: k-ε, k-ω SST RANS and an IDDES formulation. Results from each method have been compared against the full-scale ship propeller torque data of the MV Regal, a 138m long general cargo vessel. Additionally, the capability of each turbulence strategy to resolve time-dependent features of the flow, such as the bilge vortex and its effect on the boundary layer velocity fields, has been evaluated. The results from this investigation show that the IDDES based numerical model replicated the sea trials measurements with the highest degree of accuracy. Furthermore, this study confirmed that the choice of turbulence strategy has a major impact on the full-scale velocity fields in the aft region of a ship.