Crystallite thickness distributions have been calculated using the Thomson-Gibbs equation from the endotherms obtained by differential scanning calorimetry. These distributions were then compared with those obtained by transmission electron microscopy or by Raman spectroscopy (LAM). For linear polyethylene, the experimentally;measured distribution can only be reproduced by choosing appropriate combinations of the interfacial free energy and the heating rate. The distributions calculated from endotherms of random ethylene copolymers are very different from those that are obtained directly. The nature of the fusion process of random copolymers, where the free energy of the melt is not invariant with temperature, restrains the applicability of the simple Thomson-Gibbs equation.