A new correlation is presented to predict critical temperature and critical pressure of multi-component mixtures, in which binary interaction coefficients T-ij and K-ij are used to describe non-ideal contribution of binary mixtures. The binary interaction coefficients of 515 binary mixtures for critical temperature and 319 binary mixtures for critical pressure are determined. Compared with the experimental data, the average absolute relative deviations of the correction for calculating critical temperature and pressure are 1.65% and 5.21%, respectively. In order to predict critical temperature and critical pressure of the mixtures which did not list in this paper, the Chueh-Prausnitz equation can be used to calculate the binary interaction coefficients T-ij and K-ij. The specific coefficients in the equations have been given for 9 types of binary mixtures. In order to evaluate the ability of this method to predict the critical temperature and critical pressure of multi-component mixtures, 348 experimental data were collected and the corresponding values were calculated by this method. The absolute average relative deviation of the prediction values is about 1.10% for critical temperature and 6.39% for critical pressure. (C) 2016 Elsevier B.V. All rights reserved.