Previous studies observed that the single aliquot regenerative dose (SAR) protocol is problematic for the quartz violet stimulated luminescence (VSL) signal. We therefore evaluate the reliability of multiple aliquot approaches, i.e. single aliquot regeneration and added dose (SARA) and sensitivity-corrected multiple aliquot regenerative dose (SC-MAR) protocols, for the first time using fine-grained quartz samples from a Chinese loess-palaeosol sequence in Luochuan with reference ages ranging from ~25 ka to ~1400 ka. Applying the SARA protocol on one sample with an expected natural dose of ~700 Gy, yielded a SARA D_e that was consistent with independent D_e. Although this result is very encouraging, the applicability of this method is limited to samples for which the natural dose is far from signal saturation. Using the MAR protocol, all estimated VSL ages significantly underestimate the expected ages, except the two youngest samples. We show that the observed age discrepancy is not due to the thermal instability of the signal nor to the bleaching treatment in the MAR protocol, but is related to the different shapes of the natural- and laboratory-generated dose response curves (DRCs). Unlike the natural DRC, the VSL signal continues to increase at high laboratory doses in the laboratory generated DRC; suggesting that there is additional linear growth in the laboratory DRC, which does not exist in nature. The deviation between the natural- and laboratory-generated DRCs points out that the maximum dating limit of fine-grained quartz VSL on the Luochuan section under selected measurement conditions using the MAR protocol is ~250 Gy. However, it is remarkable that the natural VSL DRC saturates at about 900 Gy, which would potentially allow dating at the Luochuan section using VSL up to ca. 300 ka.