Dating sediment older than ∼150ka still remains difficult because of a lack of suitable chronological tools. Here we report on studies to determine whether measurements of quartz thermoluminescence made at constant temperature (isothermal luminescence) can be used to extend the dose limit in sediment dating. The origins and dosimetric characteristics of the quartz isothermal signal at 310∘C are investigated. It is shown that, in general, application of a single aliquot regenerative (SAR) dose protocol with this signal overestimates the expected equivalent doses in sedimentary samples, from a variety of environments and locations. This anomaly originates from a change of luminescence sensitivity, during or after the initial thermal measurement. A single aliquot regenerated and added dose (SARA) protocol suggested as the only approach known to compensate for this initial change of sensitivity was unsuccessful.