The Ozawa, Kissinger and Boswell isoconversion methods for obtaining activation energies, E a , from experiments performed at constant heating rate belong to one group of methods. It is shown that from these three methods the Kissinger method is generally the most accurate. Based on the analysis of the approximation errors made in this group of methods, a new isoconversion method is obtained, which takes the form:ln βT 1 . 8 f = -AE a k B T f + constant where A = 1.0070-1.2 10 - 5 E a (E a in kJ/mol), β is the heating rate and T f is the temperature at a fixed amount transformed. Hence, similar to Ozawa, Boswell and Kissinger's methods it is based on obtaining the slope of a logarithmic function containing the heating rate vs. 1/T. The new method is shown to be significantly more accurate than the others.