To evaluate the relationship between the exhaust gas purification performance and the catalyst pore properties related to gas diffusion, an elementary reaction model was combined with gas diffusion into catalyst pores, referred to as the pseudo-2D gas diffusion/reaction model. It was constructed for Pt/Al2O3 + CeO2 catalyst as lean NOx catalyst. The gas diffusion was described as macro pore diffusion between the catalyst particles and meso pore diffusion within the particle. The kinetic model was composed of 26 reactions of NO/CO/O2 chemistry including 17 Pt/Al2O3 catalyst reactions and 9 CeO2 reactions. Arrhenius parameters were optimized using activity measurement results from various catalysts with various pore properties, meso pore volume and diameter, macro pore volume and diameter, particle size, and washcoat thickness. Good agreement was achieved between the measured and calculated values. Moreover, as compared to the performance of other catalysts which was not used for reaction model calibration, the simulated results were in good agreement with this experimental data. Using the calibrated reaction model combined with detailed gas diffusion model, a sensitivity analysis for NOx concentration was performed to identify key diffusion parameters of the lean NOx catalyst. In summary, parameters with the highest sensitivity were catalyst particle size, washcoat thickness, CeO2 meso pore volume, and diameter. Also, catalyst macro pore volume and diameter showed no dependence. Consequently, to increase the NOx purification performance, the meso pore parameters and washcoat thickness should be increased, particle size should be decreased. With these changes, the gaseous NOx will easily move to the catalyst surface and into the active site of meso pore, resulting in better performance.