The possibility of applying the Weibull mixture model for the fitting of the non-isothermal conversion data for the combustion process of animal bone samples has been investigated. It has been found that the conversion data at the different heating rates for the investigated samples, can be successfully described by the linear combination of few Weibull distribution functions. An optimal fitting of the conversion data has been obtained by a mixture of three Weibull distribution functions. It was established that the conversion curves calculated by the proposed model, are in good agreement with the raw conversion data of the investigated bone samples. The alterations of distribution parameters in the Weibull mixture model, indicate a probably change of the kinetic nature of the particular reaction stage. A model with a continuous distribution of the apparent activation energies (E-a) can adequately represents the individually stages of the complex combustion process. The characterization of the estimated distribution curves, enable us to receive the additional information's about the kinetic behaviour of individual reaction stages, which does not attainable from the results of conventional kinetic analysis. (C) 2010 Elsevier B.V. All rights reserved.