The liquid−vapor interfacial properties of hydrocarbons and their mixtures are important factors in a wide range of industrial processes and applications. Determining these properties experimentally, however, is not only practically demanding, but many important properties, such as phase densities and compositions are not directly experimentally accessible, thus requiring the development of theoretical models. Molecular dynamics (MD) simulations, by contrast, are relatively straightforward even for the most complex of mixtures and directly provide all of the microscopic quantities for the studied systems. We have previously applied MD simulations to study the liquid−vapor equilibria of mixtures of hydrocarbons and CO 2 that are particularly relevant to hydrocarbon recovery from geologic formations. In this study, we explore in more detail the robustness of the simulation methods with respect to the choice of the model system parameters, investigate the accuracy of the simulations in determining the key quantities: system pressure and interfacial tension (IFT), and, finally, devise a simple correction for achieving a much closer agreement between the simulated and experimental quantities. We perform extensive MD simulations for three mixtures, propane/n-pentane, propane/n-hexane, and CO 2 /n-pentane, using model systems from 1000 up to 100 000 molecules, and different simulation box dimensions to test for the sensitivity to finite-size effects. The results show that changing the system size and box dimensions does not significantly impact the accuracy of the simulations. Subsequently, we examine the accuracy of the MD simulations in determining the pressure and IFT for two pure hydrocarbon systems, n-pentane and n-heptane. Finally, we propose a simple linear correction formula for the pressures and IFTs obtained from the MD simulations that closely reproduce the experimental values for single components and mixtures of hydrocarbons. Our results enable the MD simulations to provide more accurate and reliable predictions of the interfacial properties, thereby reducing the need for challenging laboratory experiments.