Development of high pressure-high temperature (RPHT) reservoirs is increasingly being pursued world wide to exploit hydrocarbon from deep formation. Due to their extreme conditions (can be up to 2000e and 16000 psia), composition of RPHT fluids can be considerably different from that of conventional reservoirs, particularly the concentration of heavy hydrocarbons in the vapour phase can be quite high. Also, at high temperature the amount of water dissolved in the reservoir hydrocarbon phase could be significant and should be taken into account in detennining the phase and volumetric properties of the fluids. In this study, a model based on equations of state (BOS) was developed to predict phase behaviour of RPHT fluids in the presence of water. The conventional mixing rule was modified by adding a non-random element in the attractive term of EOS. This modification was required to describe the interaction between non-polar (hydrocarbon) and polar (water) compounds. The developed phase behaviour model with the added term to the conventional mixing rules was evaluated for predicting the phase behaviour of hydrocarbon mixtures in presence of water. The conventional (random) interaction parameters (kij) and the non-random interaction parameter of the asymmetric term (lpi) for water-hydrocarbons were determined by matching the solubility data of hydrocarbon-water binary systems in vapour-liquid equilibrium for light and in liquid-liquid equilibrium for intermediate and heavy hydrocarbons. A method based on the Krichevsky-Kasarnovsky equation was developed to correct the effect of pressure on fugacity of the solute in the liquid phase in liquid-liquid equilibrium. The determined binary interaction parameters (BIP) were generalised by correlating them with critical properties and the molecular weight of hydrocarbons. The reliability of the model was evaluated against measured data, not used in its development, over a wide range of pressure and temperature and compared with those of leading models reported in the literature. The model could reliably predict the presence of free water phase and the effect of pressure on the liquid water phase at high temperatures. It also reliably predicted the effect of water on saturation pressure of tested synthetic reservoir fluids. However, it failed to accurately reproduce the effect of addition of water on the volumetric behaviour of the liquid hydrocarbon phase.