A new method to determine the correct Gibbs-Thomson plot and equilibrium melting temperature (T-m(0)) of polymers was proposed. The Gibbs-Thomson plot method is reliable, because the Gibbs-Thomson equation is directly derived from thermodynamical relations. In this method, the heating rate dependence of melting temperature (T.) was omitted by applying the theory of the "melting kinetics", and the effect of lamellar thickening on T. was also omitted by observing thick lamellae. A differential scanning calorimeter (DSC) was used for observation of T. as a conventional method. Transmission electron microscope (TEM) was used to observe a distribution of lamellar thickness (l). It was shown theoretically that peak temperature of melting endotherm (T.(DSC)) corresponded to averaged reciprocal l () for the case of sharp distributions of T-m and l(-1). The Gibbs-Thomson plot, T-m (DSC) vs , was carried out. A reliable Gibbs-Thomson plot and T-m(0) = 186.2 degreesC were obtained for a fraction of isotactic polypropylene (iPP) with high tacticity ([mmmm] = 99.6%, M-n = 64 x 10(3) and MwMn = 2.4). It was shown that DSC double melting endotherm corresponded to the number-distribution of l(-1), when lamellar thickening did not occur.