A complete description of fluid-phase thermodynamic properties at cryogenic conditions undoubtedly presents a serious challenge for equations of state (EOSs). This work is focused on comparing the capabilities of two Statistical Associating Fluid Theory-based EoSs, namely PC-SAFT, SAFT + Cubic, and one cubic model, the original PR EOS, in correlating and predicting various thermodynamic properties of 14 main cryogenic fluids (noble gases, light hydrocarbons, nitrogen, oxygen, carbon dioxide, carbon monoxide, and nitrous oxide). Additionally, the available low temperature vapor-liquid equilibrium (VLE) data of their binary and ternary mixtures are also evaluated in this comparative study. It has been found that, apart from the critical and near-critical data, PC-SAFT has a clear advantage in modeling the pure compound properties such the sound velocity and density (with the overall AADs% less than 7%). Nevertheless, the overestimation of the single phase properties in low temperature and high pressure regions and the poor accuracy in modeling the high pressure VLE compositions of binary mixtures are two major drawbacks of PC-SAFT. SAFT + Cubic is less accurate in predicting pure compound properties; however, having small values of binary interaction parameters, it is slightly more accurate than PC-SAFT in modeling various VLE isotherms of mixtures in a predictive manner (with the overall AADs around 3%). At the same time, PR EOS exhibits the most significant deviations from the low temperature experimental data of both pure cryogens and their mixtures. However, it yields particularly accurate predictions of isochoric heat capacities in the liquid phase. (C) 2019 Elsevier B.V. All rights reserved.