The performance of journal bearings is dependent on many factors such as design, materials, load cycle and lubricant. The behaviour of lubricants in automotive bearings is of critical importance as the size of the bearing is restricted due to engine compartment space constraints and the loadings are increased with rising power output. The traditional approach to the modelling of journal bearing lubrication has been via the “lubrication approximation” introduced by Reynolds. If the lubrication approximation is not invoked, then there is no option but to solve the full set of coupled equations (kinematic and constitutive) governing the flow of the lubricant and taking proper account of the moving parts of the geometry. Until recently, this task has proved too formidable, but with current computing power, combined with efficient and accurate numerical methods, the calculation may be attempted. In this paper a moving spectral element method is used, together with sophisticated computer visualization captured on video, to demonstrate cavitation and pressure and temperature variation in the oil film around a transiently loaded journal bearing. The effects of these variables and shear thinning on lubricant viscosity are also shown. The lubricant flow in the journal bearing is demonstrated to be extremely complex, with lubricant piezoviscous effects dominating temperature effects in the generation of oil film thickness at the position of closest approach of the journal and stator surfaces.