An investigation of the effect of mixing on gibbsite precipitator performance first requires that the term mixing be defined. This widely used term has as many definitions as there are mixing applications. Mixing has spatial aspects, e.g. dispersion, scale of segregation, inhomogeneity. Mixing also has a number of temporal aspects, e.g. the kinetics of mixing, age and residence time distributions. Physically, it can be visualised as the stretching and folding of fluid elements or as a process that randomly distributes particles. Mixing is different to shear. Although the two phenomena are related, their effects on precipitation are expected to be different. This paper describes the conceptual and practical differences between mixing and shear, and the difficulties in quantifying these terms. The manner in which mixing can effect precipitator performance is discussed in a general sense. The results of gibbsite precipitation experiments, in which the mean shear rate was kept constant and the flowpatterns in the vessel were varied, are reported. These results suggest for gibbsite precipitation in a turbulently stirred vessel that mixing has negligible effect on the product crystal size distribution. As agglomeration was a significant size enlargement mechanism in these experiments, this result suggests that agglomeration is not significantly effected by mixing. A simulation of particle-particle collisions has shown the collision frequency to be sensitive to mixing. This suggests that particle-particle collisions are not the rate-determining step in gibbsite agglomeration.