Single aliquot protocols are now widely used as a means of measuring the equivalent dose (De) in quartz and feldspar optical stimulated luminescence (OSL) dating of both heated and sedimentary materials. The most recent of these is the single-aliquot regenerative-dose (SAR) protocol, first suggested by Murray and Roberts (Radiation Measurements 29, 503–515, 1998). In this approach, each natural or regenerated dose OSL measurement is corrected for changes in sensitivity using the OSL response to a subsequent test dose (10–20% of De). If the sensitivity correction is adequate, then the corrected OSL response should be independent of prior dose and thermal/optical treatment, i.e. there should be no change in the sensitivity-corrected dose–response curve on remeasurement. Here we examine the interpretation of the sensitivity corrected growth curve as a function of dose, and the effect of changing measurement conditions (e.g. preheat temperature, size of test dose, stimulation temperature) on the estimation of De. The dependence of the dose response on prior treatment is tested explicitly, and the significance of thermal transfer discussed. It is concluded that a robust SAR protocol is now available for quartz, and that it is applicable to a wide range of heated and unheated materials.