In this paper the use of neural networks for fitting complex kinetic data is discussed. To assess the validity of the approach two different neural network architectures are compared with the traditional kinetic identification methods for two cases : the homogeneous esterification reaction between propionic anhydride and 2-butanol, catalysed by sulphuric acid, and the heterogeneous liquid-liquid toluene mononitration by mixed acid. A large set of experimental data obtained by adiabatic and heat flux calorimetry and by gas chromatography is used for the training of the neural networks. The results indicate that the neural network approach can be used to deal with the fitting of complex kinetic data to obtain an approximate reaction rate function in a limited amount of time, which can be used for design improvement or optimisation when, owing to small production levels or time constraints, it is not possible to develop a detailed kinetic analysis.