In a previous work, a predictive model for the interaction parameter k(ij) was proposed as k(ij) = k(i) - k(j). The assumed mixing factor k(i) for the pure components can be obtained by fitting to the correlated k(ij) of mixtures which have available experimental data. In this work, up to 62 binary mixtures containing HFCs and HCs were involved and their reported experimental VLE data were correlated using the PR equation of state with the vdW mixing rules. According to the correlation results, a modified differential-model was proposed for the prediction of the interaction parameter k(ij) with the introduction of the acentric factors and the critical properties of the components. The values of mixing factors k(i) in the modified model were obtained by the least square fitting of the correlated 62 k(ij) data for 15 HFCs and HCs, namely HFC23, HFC32, HFC125, HFC143a, HFC134a, HFC134, HFC152a, HFC161, HFC227ea, HFC236fa, HFC236ea, HFC245fa, propane, isobutane, and n-butane. 105 predicted k(ij) for mixtures containing these HFCs and HCs were given. The predicted results by using the modified model show a good agreement compared to the correlation results. The overall average absolute deviations of pressures and vapor phase compositions of the predicted results are 1.77% and 0.0086, respectively. (C) 2012 Elsevier B.V. All rights reserved.