When electron spin resonance (ESR) is applied to sedimentary quartz, dealing with the poor bleachability of the signals is particularly challenging. In this study, we used both the single-grain optically stimulated luminescence (OSL) and the single aliquot ESR dating of quartz from deep sand deposits preserving a Stone Age archaeological sequence to combine the advantages of the two methods: good bleaching behaviour and extended age range. Using the youngest samples at each sampling site we were able to calculate the mean ESR residual age from the difference between the OSL ages and the apparent ESR ages. Focusing mainly on the single aliquot regenerative dose (SAR) protocol here, we were able to calculate the mean ESR residual age for the Ti and Al centres, including the non-bleachable signal component for the latter. For the NP site, residual ages of 209 ± 13 ka and 695 ± 23 ka were calculated for the two centres, whereas for the ZS site 268 ± 39 ka and 742 ± 118 ka were determined. These residual ages are significant and cannot be neglected. Thus, the residual age was subtracted from the apparent ESR ages. The validity of the residual subtraction method was tested through a comparison of the oldest OSL age from each site with the residual subtracted ESR age. For both NP and ZS sites, the residual subtracted Ti and Al ages were consistent with the OSL age within 2-σ uncertainty, and therefore confirm the robustness of the subtraction method. Within the NP sequence, we were able to locate the end of the Early Stone Age at 590 ± 86 ka, and this provides a maximum age for the transition to the Middle Stone Age in this part of south-central Africa.