In recent years, there has been an increasing interest of the petrochemical industry in modeling of the partitioning of production chemicals e.g. gas hydrate inhibitors, corrosion inhibitors, solvents etc. between the crude oil and water. This requires basically a thermodynamic model either in terms of an activity coefficient model or an equation of state. Our target in this thesis is to review and develop such models capable of describing qualitatively as well as quantitatively phase equilibria in multicomponent multiphase systems containing non-polar, polar, and associating compounds. The background and main targets for this thesis are presented in Chapter 1. In Chapter 2, a comprehensive review of the application of group contribution (GC) models such as various forms of UNIFAC and the so-called AFC (Atom and Fragment Contributions) correlation model for Pow (octanol-water partition coefficient) calculations has been carried out. UNIFAC is an activity coefficient model while AFC is a model specifically developed for Pow calculations. Five different versions of UNIFAC and the AFC correlation model have been compared with each other and with experimental data. The range of applicability of the GC models to Pow is discussed, and general conclusions are obtained. A thorough analysis of the models was conducted including residual plots and numerical and graphical comparisons. We conclude that the group-contribution concept has possibly exhausted its applicability to account for highly asymmetric systems, especially for aqueous solutions with complex poly-functional chemicals. In Chapter 3, liquid-liquid equilibrium data for 7 binary glycol-hydrocarbon systems have been measured in the temperature range 32 °C to 80 °C and pressure equal to 1 bar. The measured systems are monoethylene glycol + heptane, methylcyclohexane, hexane, propylene glycol + heptane, diethylene glycol + heptane, triethylene glycol + heptane, and tetraethylene glycol + heptane. The data obtained were correlated with the NRTL model and two different versions of the UNIQUAC equation. The NRTL model and one of the UNIQUAC equations (UQ 4) have a linear temperature-dependent interaction parameter term, while the other UNIQUAC equation (UQ 2) has an interaction parameter that is independent of the temperature. There was a fairly good agreement between the experimental data and the two temperature dependent models with an average deviation in the composition for both phases of 3 % for both NRTL and UQ 4 while deviation is 15 % for UQ 2. These results indicate the necessity of using the linearly dependent interaction parameters. The CPA equation of state is a thermodynamic model, which combines the well–known cubic SRK equation of state and the association term proposed by Wertheim, typically employed in models like the various variations of SAFT. CPA has been shown in the past to be a successful model for phase equilibria calculations for systems containing water, hydrocarbons and alcohols. In Chapter 4, CPA is applied for the first time to liquid-liquid equilibria for systems containing glycols and hydrocarbons. It is shown that excellent correlation is achieved with solely a single interaction parameter per binary system. The correlation procedure as well as the nature of the experimental data play a crucial role in the parameter estimation and they are thus extensively discussed. In Chapter 5, the application of the CPA equation of state is extended to mixtures containing cross-associating compounds such as glycols and water. In this case, combining rules are required in the association term of CPA for the cross-association energy and volume parameters. Different types of such combining rules have been suggested over the past years for association models such as SAFT. These are tested in this work for CPA in terms of their correlation and prediction capabilities for vapor-liquid equilibria of glycol-water systems. Comparisons with SRK are also provided. It was found, that the arithmetic mean combining rule for the cross-association energy parameter and the geometric mean for the cross-association volume parameter provide overall the best results for cross-associating systems containing glycols and water. Moreover, preliminary results show that the CPA model can be used to predict multi-component, multiphase equilibria for glycol/water/hydrocarbon mixtures based solely on binary interaction parameters. In Chapter 6, conclusions and suggestions for future work are presented.