This paper investigates the effectiveness of employing the traditional Leverett approach for describing the capillary-induced flow in thin-film fuel cell diffusion media (DM) with mixed wettability. A one-dimensional steady-state analytical model is developed to analyze the capillary transport of liquid water through the thin-film DM. A capillary pressure-liquid saturation relation is derived based on the experimental data reported by Gostick et al. [J.T. Gostick, M.W. Fowler, M.A. loannidis, M.D. Pritzker, Y.M. Volfkovich, A. Sakars, J. Power Sources, 156 (2006) 375]. The empirical fit is applicable to the fuel cell DMs tested by reference [J.T. Gostick, M.W. Fowler, M.A. loannidis, M.D. Pritzker, Y.M. Volfkovich, A. Sakars, J. Power Sources, 156 (2006) 375] under a limited set of conditions, providing a means to evaluate the effectiveness of the traditional Leverett approach. Furthermore, a new characteristic relative permeability correlation appropriate for the tested DMs under the given conditions was obtained by fitting the experimental capillary pressure data [J.T. Gostick, M.W. Fowler, M.A. loannidis, M.D. Pritzker, YM. Volfkovich, A. Sakars, J. Power Sources, 156 (2006) 375] into four well-established empirical models. The empirically-derived relationships are then integrated into an analytical model framework in order to compare the liquid saturation profiles predicted by both approaches. The results show that use of the standard Leverett approach equipped with Leverett J-function can lead to significant errors; therefore, an extension of this approach appropriate for fuel cell DM with mixed wettability is needed for reliable model predictions. Finally, a sensitivity analysis is performed to assess the relative significance of the various input parameters on the predicted saturation profiles. (C) 2007 Elsevier B.V. All rights reserved.