The aim of this study was to compare different models, either originating from literature or originally proposed in this study, for the interpretation of the melting behavior of polymers. In particular, these models, tested with a linear low-density polyethylene widely used in rotational molding, are suitable for coupling with energy balances in the study of polymer processing. We obtained the experimental data from differential scanning calorimetry (DSC) dynamic scans, assuming that the endothermic flux was related to the rate of melting of the polymer. The studied models Were able to predict the broad melting temperature range typically observed during polymer melting with either a statistical or a kinetic approach. The two different approaches were compared with experimental DSC data. The analysis of model performances with complex thermal programs showed that the statistical approach could provide a more realistic representation of polymer melting. These models were particularly suitable in rotational molding, where the lack of any flow and, hence, of any crystalline orientation leads to a degree of melting determined by the actual temperature of the polymer. (C) 2003 Wiley Periodicals.