Monte Carlo (MC) simulations were performed to understand the statistical aspects of the local density distributions of sub- and supercritical CO2 along the VLE line and at T-r = 1.02 which were obtained from the Voronoi tessellation. From the statistical analysis, the local density distribution of CO2 (v) and CO2 (I) along the VLE line was assumed to be represented by an Inverse Gamma distribution and a Normal distribution respectively. An Inverse Gamma-Normal mixture model was then proposed as a statistical model for the representation of the local density distributions of CO2 (SC). For the density distribution of CO2 (SC) at T-r = 1.02, the mixture model was well fitted to the distribution The mixture model could also find out the gas-to-liquid transition point located at rho(r) congruent to 0.92. In addition, the mean local densities were correlated to the conventional concept of the effective local density. (C) 2016 Elsevier B.V. All rights reserved.