In our previous paper we applied the Song and Mason (SM) equation of state (EOS) to refrigerant fluids, using the speed of sound data. Here, we predict the equation of state for mixtures of refrigerants 1,1,1,2-tetrafluoroethane (R134a), 1,1-difluoroethane (R152a), difluoromethane (R32), pentafluoroethane (R125), and 1,1,1-trifluoroethane (R143a). The temperature-dependent parameter second virial coefficient, B(P, of pure refrigerants that are necessary for SM EOS are found using the two-parameter corresponding states correlation obtained from the speed of sound data analysis and two constants: the enthalpy of vaporization DeltaH(vap) and the molar density Pnb, both at the normal boiling point. Other temperature-dependent quantities, including the correction factor (alpha(T), and van der Waals covolume b(T) are obtained from the second virial coefficients by the scaling rules. The cross parameters B-12(T), alpha(12)(T), and b(12)(T) as required by the EOS for mixtures are determined by the use of a simple combining rule. This EOS is applied to six binary refrigerant mixtures, including R125 + R134a, R32 + R125, R125 + R143a, R32 + R134a, R134a + R152a, and R32 + R143a, and two ternary mixtures, R32 + R125 + R134a and R32 + R125 + R143a. The results show that the molar liquid and gas densities of the refrigerant mixtures can be predicted to within 5% over a wide range of temperatures and pressures. Also, the molar liquidlike densities of R32 + R134a and R152a + R134a mixtures obtained from the SM EOS has been compared with those calculated from the original Peng-Robinson (PR) and Nasrifar-Moshfeghian (NM) equations of state. Our results are in favor of the preference of the SM EOS over two other equations of state.