Three different lumped kinetic models for a wet oxidation process were used in this study to describe an originally complicated wet oxidation reaction system. The model parameters, the reaction orders for oxygen, frequency factors, and activation energies for the model reactions, were estimated by comparing measured data to predictions made using the model. Sensitivity analysis, a study on the cross-correlation and identification of the parameters, was carried out systematically during the parameter estimation. The models, which are presented together with the estimated parameters, explained the measured data to a good degree of agreement. This study shows that with proper parametrization the obtained parameters are well-identified and do not significantly correlate with each other. This is important in order to obtain meaningful parameters that are as reliable as possible. The parameter estimation and sensitivity analysis showed that the more complicated models, which include the biodegradability changes, were statistically more reliable and their prediction properties were better. It was also demonstrated that the wet oxidation reaction rates are dependent on the oxygen concentration in water.