The transformation kinetics from glass to crystalline of two clearly separated crystallization peaks of Si12.5Te87.5 chalcogenide glass were studied using DSC technique. The kinetic parameters of these two separated crystallization peaks of the present chalcogenide glass under non-isothermal conditions are analyzed by the theoretical method developed (TMD) and the isoconversional (model-free) with the help of model-fitting method for the heating rates range of 5-90 K min(-1). The average values of the effective activation energies, E-eff., are 133.27 +/- 5.29 kJ mol(-1) for the first crystallization peak (FCP) and 198.99 +/- 4.81 kJ mol(-1) for the second crystallization peak (SCP). The dependence of the transformed fraction, alpha(p), corresponding to the maximum transformation rate, on impingement exponent, gamma i, indicated that the mode of impingement is due to "anisotropic growth". The average value of kinetic exponent, n(p), calculated using TMD method is 1.53 +/- 0.07 for FCP, while, the average value of n(p) for SCP is 5.55 +/- 0.52. The unexpected high value of the kinetic exponent for SCP could be explained as both nucleation frequency and crystal growth rate depend on the time as a power law. The effective activation energy of crystallization, E-eff.(alpha), using the isoconversional method, shows a little variation throughout the entire interval of transformations for FCP compared to that for SCP. Whereas, the dependence of E-eff.(alpha) for SCP shows a strong dependence on the degrees of transformation and temperature. The reaction model that may describe transformation process of Si12.5Te87.5 chalcogenide glass for FCP is Avrami-Erofeev model (g(alpha)=[-In(1 - alpha)](1/n)) with n equal to 1.5 for all heating rates used (5-90 K min(-1)). On the other hand, the reaction model that may describe transformation process of SCP is a power law model g(alpha) = alpha(1/xi) with xi = 0.25 for heating rate range of 5-35 K min(-1) and xi = 0.50 for the heating rate range of 40-90 K min(-1). Finally, the obtained results of transformation kinetic parameters using TMD analysis are in good agreement with that obtained according to the isoconversional one under non-isothermal regime. These results indicate that each method confirms and complements the other, and using the two methods for evaluating the transformation kinetic parameters to describe the glass-crystalline transformation process of Si1.25Te87.5 chalcogenide glass is helpful. (C) 2011 Elsevier B.V. All rights reserved.