There is currently considerable debate concerning the most appropriate rheological model to describe the behaviour of lubricant films in rolling-sliding, elastohydrodynamic contacts. This is an important issue since an accurate model is required in to predict friction in such contacts. This paper reviews the origins of this debate, which primarily concerns a divergence of views between researchers using high pressure, high shear rate viscometry and those concerned with the measurement and analysis of elastohydrodynamic friction; the former advocate a Carreau-based shear stress/strain rate model while the latter generally favour an Eyring-based one. The crucial importance of accounting for shear heating effects in analysing both viscometric and friction data is discussed. The main criticisms levied by advocates of a Carreau-based model against Eyring's model are discussed in some detail. Finally the ability of both types of rheological model to fit elastohydrodynamic friction measurements for a quite simple, well-defined base fluid is tested, using previously-measured pressure-viscosity behaviour for the fluid. Both models appear to fit the experimental data over a wide temperature range quite well, though fit of the Eyring model appears slightly closer than that of the Carreau-Yasuda model. Friction data from a wider range of welldefined fluid types is needed to identify categorically the most appropriate model to describe elastohydrodynamic friction.