A computational method based on molecular dynamics (MD) simulation is developed to predict the interaction parameters in local composition (LC) models such as Wilson, nonrandom two-liquid (NRTL), and universal quasichemical activity coefficient (UNIQUAC) applicable in vapor-liquid equilibrium calculations so that LC models are validated. The five binary mixtures of water-acetonitrile, water-isopropanol, methanol-chloroform, acetone-cyclohexane, and meta xylene-benzene were simulated. The MD simulations are performed using the condensed phase optimized molecular potential for atomistic simulation studies force field and all ranges of radial distribution function (RDF) are considered. In addition, the interaction parameters are determined by optimization through experimental data and are compared with the MD results. The interaction parameters are calculated through RDF where the effective radius is investigated by experimental data. The present results demonstrate that interaction parameters are composition dependent which are best fitted by a third-order polynomial relation. Based on the deviation in the activity coefficients, the results of the UNIQUAC model are more accurate than Wilson and NRTL models. (C) 2015 American Institute of Chemical Engineers