Cubic equations (EoSs) of state are successfully used in petroleum and natural gas industry. In order to extend these equations to mixtures, van der Waals mixing rules with Lorentz-Berthelot (LB) combining rules are often employed; however, the accuracies of these EoSs in predicting the densities of hydrocarbon mixtures are not adequate. The main objective of this study was comparing 13 EoSs coupled with 10 combining rules in predicting the densities of hydrocarbon mixtures. Binary and ternary liquid mixtures, LNG mixtures and synthetic natural gas mixtures comprising 752 data points were collected and used in this study. Results revealed that for predicting the liquid densities of binary and ternary mixtures, the Schmidt and Wenzel (SW) EoS coupled with Hudson-McCoubrey (HMC) or LB combining rules are the best among the others. The SW EoS coupled with the LB combining rules were also the best in predicting the densities of the LNG mixtures. Additionally, the LB combining rules are the best in predicting natural gas mixtures densities using the Patel and Teja (PT), SW and Patel-Teja-Valderrama (PTV) EoSs. In general, it was found that the Redlich-Kwong (RK) family EoSs, SW and Trebble-Bishnoi-Salim (TBS) EoSs are best coupled with the LB combining rules. However, the Peng and Robinson (PR) family EoSs with Halgren (HHG) combining rules were in better agreement with experimental data. The Waldman-Hagler (WH) combining rules lacked predictability when coupled with the most EoSs.