New analytical techniques have opened up the possibility of addressing rates of soil processes quantitatively. Here, we present the results of an investigation into the use of single-grain optically stimulated luminescence (OSL) dating to derive rates of soil mixing in the top 50 cm of soil profiles from two toposequences situated in the Werrikimbe National Park in Australia. Of 500 single grains analysed from each sampled depth increment, less than 25% provided a finite age, with the rest of the grains either non-responsive or dose-saturated. This proportion of finite-age grains tended to decrease with soil depth. Median ages of quartz grains increased down the soil profile, with topsoil ages of up to 500 years and subsoil ages of up to 5000 years. Few 'younger' grains were found deeper in the profile and few 'older' grains near the soil surface. These trends suggest that pedoturbation is resulting in vertical transport of grains through the profile, but that there is a distribution of transport distances, with a poor probability of large transport distances from surface to subsoil or vice versa compared with a more frequent movement of grains to and from the surface in the uppermost 10-35 cm. The calculation of a single age for each soil horizon was unachievable as each horizon contained a heterogeneous mixture of grains with varying histories of transport to and from the soil surface. Soil mixing was confirmed along both toposequences studied. However, the occurrence of minor mixing rates did not lead to a homogenization of the topsoil and adjacent horizons. We postulated that mixing velocities were mostly related to flora at our study site. Vertical soil mixing rates of 0.5 and 0.2 mm year-1 were calculated from the distribution of finite single-grain ages.