The study of quartz grains (diameter 90-125 mu m) from a modern dune sand from Australia shows a lack of perfect proportionality between the regenerated OSL (L) signal and the subsequent test dose OSL (T) signal when using a single-aliquot regenerative dose (SAR) procedure with a fixed 15 Gy dose repeated five times. This lack of perfect proportionality results in both a broad D, distribution and systematic offset for the doses measured both with and without a preheat step at the end of each cycle of the SAR procedure. The problems caused by the lack of proportionality are overcome by a new method of handling the data generated by use of the SAR protocol; the OSL intensity is sensitivity-corrected using the slope of a plot of L versus T obtained from a number of aliquots, rather than the ratio of LIT obtained for individual aliquots. This approach is validated using laboratory irradiated quartz with given doses ranging from 5 Gy to 60 Gy; for the larger doses, correct values can be recovered using the new data analysis procedure is applied, but not when using the ratio of LIT obtained using conventional SAR analysis.