Exergy analysis has been applied to desalination membrane processes in an effort to characterise energy consumption and to optimise energy efficiency. Several models have been used to this end in the literature. One assumption that is common in these analyses is that of ideal solution behavior. However, seawater and other aqueous solutions of interest do not behave ideally. Indeed, even when ideal behavior is not assumed, there are several approaches to calculate these activity values, which are typically a function of the molality and ionic strength of the electrolytic solution. What is not clear from the published literature is the impact that the choice of activity calculation model has on the exergy analysis results. The objective of this research was to undertake the exergy analysis of a seawater membrane desalination plant using the Szargut chemical exergy approach and to compare the activity calculation approaches. The chemical exergy of the seawater was calculated using several activity coefficient modelling approaches including, (a) ideal mixture model, (b) the Debye-Huckel limiting law, (c) the Davies model, and finally, (d) the Pitzer model, which is more appropriate for higher ionic strength solutions such as seawater. The results showed considerable differences in the chemical exergy rates and the magnitude of chemical exergy destruction rates calculated using the various models. For example, there were percentage differences of 61.8% and 44.7% between the magnitude of chemical exergy destruction rates calculated using the Pitzer model when compared with the Debye-Huckel limiting law for the nanofiltration and reverse osmosis processes respectively.