SCR (selective catalytic reduction) is one of the main after-treatment systems currently to control engine NOx emission, and its structure parameters affect its performance and cost directly. In this study, the structural parameters of the SCR reactor are optimized by considering the coupling relationship between each structural parameter by using the RSM (Response Surface Methodology). Finally, the new reactor with the optimized parameters was simulated to double check its NOx reduction ability. A 1D model of SCR system is constructed using AVL BOOST software. The influence of structure parameters, such as catalyst cross-section area, catalyst length, substrate wall thickness, washcoat thickness, substrate cell density (CPSI), have been taken into consideration to study their effects on the SCR performance. Using BBD (Box Behnken Design) experimental design method, the tests of every factor under different levels are carried out by Design Expert software. The corresponding regression coefficients are obtained by analysis. A regression equation, in which the SCR reactor volume is the objective function, the NOx conversion rate of reactor and pressure drop are the constraint condition, was established. Based on the response analysis and optimization, a set of those mentioned parameters of the SCR reactor was obtained. Optimized structure parameters gotten from RSM are validated by using AVL boost. A series of simulation work was carried out to validate the conversion efficiency of SCR reactor. Simulation results show that volume of SCR reactor is significantly reduced compared with original SCR reactor while keeping high NOx conversion efficiency and low pressure drop. Therefore, with simulation platform, structural parameters of the SCR reactor can be effectively optimized and achieve the goal of downsizing.