Thermo-gravimetric analysis (TGA) was performed on grape seeds, skins, stalks, marc, vine-branches, grape seed oil and grape seeds depleted of their oil. The TGA data was modeled through Gaussian, logistic and Miura-Maki distributed activation energy models (DAEMs) and a simpler two-parameter model. All DAEMs allowed an accurate prediction of the TGA data; however, the Miura-Maki model could not account for the complete range of conversion for some substrates, while the Gaussian and logistic DAEMs suffered from the interrelation between the pre-exponential factor k(0) and the mean activation energy E(0) - an obstacle that can be overcome by fixing the value of k(0) a priori. The results confirmed the capabilities of DAEMs but also highlighted some drawbacks in their application to certain thermodegradation experimental data. (C) 2011 Elsevier Ltd. All rights reserved.