In this work a novel modelling approach based on Computational Fluid Dynamics (CFD) for the prediction of the gas separation process in a Pd-Ag membrane module for H-2 purification is presented. With this approach, the pressure and velocity flow fields of the gas mixture and the species concentration distribution in the selected three-dimensional domain are simultaneously and numerically computed by solving the continuity, momentum and species transport equations, including a gas-through-gas diffusion term derived from the Stefan-Maxwell formulation. As a result, the H-2 permeation calculations depend on the local determination of the mass transfer resistances offered by the gas phase and by the membrane, which is modelled as a permeable surface of known characteristics. The applicability of the model to properly predict the separation process under a wide range of pressure, feed flow rate, temperature and gas mixtures composition is assessed through a strict comparison with experimental data. The influence of inhibitor species on the module performance, that is obtained by implementing in the CFD model a suitable literature correlation, is also discussed. (C) 2009 Elsevier B.V. All rights reserved.