Abstract Recent studies have coupled blade element momentum (BEM) theory with the Reynolds Averaged Navier–Stokes equations in computational fluid dynamics (CFD) software, as the BEM-CFD method to analyse the flows in marine current turbines is with much less computational resources. The accuracy of the BEM-CFD calculation was evaluated by analysing the performance and flow field characteristics of an isolated horizontal axis marine current turbine with comparisons to a full rotor geometry simulation and experimental data. The comparisons show that the full rotor geometry simulation gives good predictions near the optimal conditions (TSR = 5–7), but is less accurate for off-design conditions. The BEM-CFD results, which are based on two-dimensional hydrofoil theory, are evaluated using the experimental and numerical lift and drag coefficients. It shows that the two-dimensional lift and drag coefficients had significant effects on the BEM-CFD predictions. Overall, the BEM-CFD based on the numerical hydrofoil data can accurately predict the thrust, but generally overestimates the power. The influence of the lift and drag terms on the BEM-CFD predictions suggest that more reasonable 2D predictions for hydrofoils and the 3D effects should be considered to improve the BEM-CFD accuracy. BEM-CFD can reasonably reflect the circumferential averaged velocity characteristics near the rotor for the optimal condition (TSR = 6) and gets symmetrical features in the wake, but it cannot predict the detailed flow features caused by the finite number of blades due to the limitations of the BEM-CFD method.