In this paper, we present a novel approach to predict the location of azeotropes for binary mixtures by two methods: from the experimental data and the thermodynamic model. The model composed of the Peng-Robinson equation of state, the Mathias-Copeman alpha function, the Wong-Sandler mixing rules involving the NRTL model. The binary mixtures of refrigerants selected are: propane (R290) + 1,1,1,2-tetrafluoroethane (R134a) [1], propane (R290) + difluoromethane (R32) [2] and hexafluoroethane (R116) + ethane (R170) [3], hexafluoroethane (R116) + carbon dioxide (R744) [4] and hexafluoroethane (R116) + propane (R290) [5], to be favorable to the environment with a null ODP (ozone depletion potential) and a low GWP (global warming). The results prove that there is an agreement between the predicted values and the experimental data and the relative error does not exceed 2.76% for the molar fraction and 3.23% for the pressure. The presented methods are able to predict the azeotropic position and the performances of the models change from one mixture to another. (C) 2014 Published by Elsevier B.V.