A heterogeneous one-dimensional model is used to simulate a three way catalytic converter (TWC). The mathematical model consists of mass and energy balance equations in the gas and solid phases and accounts for 15 heterogeneous chemical reactions. Langmuir-Hinshelwood type kinetics is used to represent most of the reaction rates. In this work, the kinetic parameters of the various reactions occurring in a TWC are estimated using vehicle data based on a genetic-algorithm-based optimization. The method uses the test data to estimate the reaction kinetic parameters by minimizing the deviation between the model predictions and experimental observations. It is found that the objective function based on the cumulative mass of a species that leaves the reactor until a certain time instant is a preferred performance measure for the optimization. The sensitivity of the exit concentration of various species to various kinetic parameters was found. This is used to identify the subset of reactions which have a significant effect on the various species concentrations in the effluent stream and is used to determine the set of kinetic parameters for optimizing the reactor performance. Single parameter and multiple parameters tuning using genetic algorithm have been demonstrated successfully for a TWC application.