Novel experimental and computational methods to investigate diffusion in catalytic coatings are presented and demonstrated with Pt/gamma-Al2O3 oxidation catalyst. The catalytic layer with defined particle and pore size distributions is coated on metal plates and subsequently overlaid by an inert layer acting as an additional diffusion resistance. The samples are tested in a lab reactor for CO oxidation and diffusion-limited regime is reached above the light-off temperature. Comparisons of the results obtained with the samples coated by active layer only and the samples additionally coated by an inert layer reveal the extent of diffusion limitations. The computational part is based on digital reconstruction of the porous layer as a 3D matrix; this is achieved using macroporosity obtained from SEM cross-section images, and measured particle size distributions. Reaction and diffusion are then simulated within a small layer section and the spatially averaged results are employed in the full-scale model of the entire reactor. The simulated light-off curves are in good agreement with the experimental data. Depending on the actual gamma-Al2O3 particle size distribution, the predicted effective diffusivities of CO at 298 K are 2.6 x 10(-6) m(2) s(-1) and 4.2 x 10(-6) m(2) s(-1) (for gamma-Al2O3 particles d(90) = 7 mu m and d(90) = 22 mu m, respectively), whereas the classical random pore model predicts approximately 25-45% lower diffusivities. (C) 2012 Elsevier B.V. All rights reserved.