This paper describes and discusses laboratory experiments showing that reliable determinations of viscosity usually cannot be made on high viscosity samples of heavy oil or tar sand bitumen that have been solvent extracted from core or cuttings samples with toluene or otherwise contaminated or diluted with this solvent. The toluene solvent cannot be quantitatively removed without damaging the bitumen sample because of the concomitant evaporation of volatile materials that act as natural solvents in the bitumen. Only small amounts of natural or contaminant solvent are sufficient to effectively control the sample viscosity. Viscosity values determined on residual bitumen after solvent removal may be either lower or higher than the nominally correct value depending on whether excessive residual toluene solvent remains in the sample or excessive volatile material has been removed from the sample during the evaporation. On the other hand, the impact of solvent on the density (i.e., API gravity) of a sample is linear and there are several advantages to using solvent plus bitumen mixtures to determine density for high viscosity samples.