Faults are represented conventionally in production flow simulation models using transmissibility multipliers which capture the single phase, but not the two phase, fault-rock properties. Available data indicate that fault-rocks have similar two phase properties to sediments of the same permeability, hence existing methods can be applied to estimate two phase fault-rock properties from their intrinsic permeabilities. Two methods of representing the two phase fault-rock properties implicitly in the flow simulator are compared, using one-dimensional numerical flow models containing water-wet faults with imbibition capillary pressure curves. The method which is the closer two phase analogue of the single phase transmissibility multiplier is inappropriate, as the implementation is unreasonably unwieldy. A simpler implementation is to derive pseudo-relative permeability functions including the fault-rock properties in the upstream grid block; these properties are then incorporated directly in the simulator. Relative transmissibility multiplier functions can be back-calculated from the pseudo-relative permeability functions, and indicate how closely the single phase multiplier approximates two phase flow through the fault. Implementation in a 3D model with complex fault juxtapositions validates the approach, and a practical workflow for the routine inclusion of two phase fault-rock properties in conventional faulted flow simulation models is outlined.