Abstract Crystallization is a thermally activated process in non-crystalline and amorphous solids. The kinetics of the solid state phase transformations can be studied using thermal analysis techniques such as differential scanning calorimetry (DSC). For the kinetic analysis of the crystallization process under non-isothermal conditions, the choice of a reliable method is very important. The methods for the analysis of non-isothermal data are, in general, derived by extending the formalism developed for isothermal conditions. Most methods for the kinetic analysis of crystallization processes rely on the isokinetic hypothesis to separate the kinetics of the transformation from its dependence on temperature. It is assumed that the transformation rate can be described by a differential equation separable in α (transformed fraction) and T (temperature), i.e., for continuous heating regime, (dα/dT)=(1/β)k(T)f(α), where β is the heating rate, k(T) is the rate constant, and f(α) is the kinetic function (reaction model). The crystallization kinetics of glassy Co66Si16B12Fe4Mo2 have been studied with DSC and analyzed using non-isothermal theoretical expressions. The Avrami exponent, (n), frequency factor, (A), and activation energy, (E), of crystallization are evaluated using Matusita and Sakka (MS) and modified Kissinger equations. Besides, isoconversional kinetic analysis has been applied to DSC data for the determination of these different kinetic parameters. The isoconversional methods calculate Eα values at progressive degrees of conversion, α, without modelistic assumptions, and hence, this approach takes care of the variation of kinetic parameters with the fraction crystallized. The activation energy has been determined using both linear integral and differential isoconversional methods and also by the non-linear isoconversional method suggested by Vyazovkin and Wight. These methods are found to give consistent results for E. Furthermore, a comparison has been made among various kinetic parameters obtained using different approaches to investigate the relative applicability and usefulness of the proposed methods.