Horizontal axis marine current turbine is a viable device which can harness kinetic energy from ocean currents. It is the closest concept to be commercialised among other marine turbines. Literature shows that computational fluid dynamics (CFD) models can accurately simulate turbine performance provided appropriate numerical techniques are employed. In this paper, the influence of different numerical approaches on the performance prediction of a two bladed turbine model was assessed by towing tank results from the USNA. Two turbulence models of k-ω SST and BSL EARSM as well as three boundary layer modeling techniques, including wall function, near wall region and transitional Gamma-Theta model, were compared. The effects of using steady state or transient solution methods by applying moving reference frame (MRF) and sliding mesh were investigated. Single blade simulation instead of whole turbine model was also evaluated together with the Reynold number effect. Although Transient solution with sliding mesh method offers a simulation closer to the real condition of turbine operation with accurate results, steady state MRF provides reasonable results while saving a significant computational time as well. Therefore, authors recommend utilising steady MRF simulation of whole turbine model using k-ω SST with wall-function model for performance prediction of horizontal axis marine current turbines in a balance between simulation time and results accuracy.