Paper 49-4 As part of an evaluation of soil production functions (Wilkinson & Humphreys 2005, Wilkinson et al. 2005) we examined in detail a shallow (c. 80 cm), undisturbed forest soil on a sandstone plateau in south-eastern Australia. Soil production rates inferred from insitu cosmogenic nuclides (TCN) are of c. 10-20 m/Ma using 10Be, which are consistent with regional rates of <23 m/Ma of plateau lowering based on paleosurface reconstruction using Miocene basalts. In general terms this means that the soil profile could develop in c. 30-60 ka allowing for changes in bulk density. Large aliquot OSL dating of quartz sand showed a steady increase in age from 3.5 ka at 10 cm to 7.7 ka at 30 cm above a dispersed pebbly stone layer. Within the stone layer the age increased to 32 ka at 40 cm and 154 ka at 60 cm. This soil has therefore experienced a least one full glacial cycle i.e. from the glacial maximum of Stage 6 through to the Holocene. Whether a soil with this age range is regarded as a paleosol or not is a moot point. However, it is useful to consider whether or not other features in the soil, other than the age of quartz sand, attest to conditions spaning interglacial to maximum glacial conditions. The study soil is a regosol or dystrustept which is common to the sandstone plateau terrain of the Sydney Basin. Not knowing the age spread most pedologists would regard it as decidedly uninteresting: it is shallow, light olive brown to brownish yellow with sand to sandy loam textures and an earthy fabric giving an Oi, A(0-5 cm), E (5-8 cm), Bw (8-28 cm), Cox (28-55 cm), Cr (55-86 cm) profile with the pebbles concentrated in Cox, i.e. it has the hallmarks of a residual soil with the Cr appearing as saprolite. Detailed fabric and particle composition analysis focussed on assessing the distribution of major constituents that might provide clues to genesis. Open and infilled tubules (pedotubules) occur throughout but decrease exponentially from the E layer. Ants and earthworms are the main bioturbators but termites and cicadas are present too. The remnants of bioturbation, maculae, dominates the s-matrix in the A-Bw horizons. This matches a trend in charcoal in which only traces extend beyond the Bw. These trends, in conjunction with Holocene average ages, indicate effective bioturbation in the upper 30 cm of the soil. The reduced amount of pebbles in this zone may therefore reflect this active bioturbation because the pebbles are too large to be moved by the common bioturbators. Alternatively, the stony layer (Cox) may represent a lag deposit. Support for the latter does not come from the soil itself but from a consideration of steady state conditions in interpreting the TCN data (Wilkinson et al. 2005) where it is argued that the soil mantle was c. 30 cm shallower prior to 15 ka to account for soil production trends. The lessons from this exercise are as follows: (i) reasonably simple appearing soils may have a complex history – in this case spanning a complete glacial cycle; (ii) there may be no obvious morphological signature of this history; and (iii) a combination of OSL dating and TCN offers a way of examining residual sites that so far have been largely ignored in a paleopedology context.