At Bau de l’Aubesier, Provence, France, Mousterian artefacts and human teeth occur in inhomogeneous (“lumpy”) sedimentary deposits that include bone and tooth fragments. Due to uranium (U) uptake by the bones and teeth, the external dose rate experienced by the teeth has changed over time. Such changes must be considered when calculating thermoluminescence (TL) ages for burnt flint artefacts or electron spin resonance (ESR) ages for the teeth therein. Here, we assess the external dose rates using three methods: in situ TL dosimetry gives the modern dose rate, whereas volumetrically averaged geochemical analysis is used to estimate time-averaged dose rates that consider U uptake by U-rich sedimentary components (Blackwell and Blickstein, 2000. Quarternary International 68–71, 329–343). In ESR isochron analysis (Blackwell and Schwarcz, 1993. Applied Radiation and Isotopes 44, 243–252), multiple subsamples with variable U concentrations measure the external dose rate experienced by the tooth regardless of reworking, but do depend on the uptake model assumed. The methods each suggest significantly different dose rates. TL dosimetry for two combustion horizons measured external γ dose rates at 0.143±0.012 mGray/y in Layer IV and 0.184±0.016 mGray/y in Layer H. Geochemical analysis when volumetrically averaged to account for the influence from different sedimentary components indicates external γ dose rates ranging from as low as 0.094±0.024 mGray/y assuming RU for U-rich sedimentary components to a maximum of 0.251±0.048 mGray/y assuming EU for the U-rich components. Isochron analyses suggest more variable dose rates, ranging from a minimum of 0.009±0.040 mGray/y (LU) in the lower layers to a high of 1.103±0.160 mGray/y (EU) in Layer H1. Three isochrons indicated that the teeth had experienced U leaching or a second uptake event which would affect the accuracy of the dose rate determinations for those teeth. Other teeth may have experienced similar processes. Quaternary researchers who want reliable dates in the age ranges for ESR or TL must carefully consider both the nature of the sediment, especially U-rich bone fragments, and its geometric arrangement around the samples to be dated. Corresponding results using more than one analytic method may be required for a reasonable confidence level.