In this study, we report standard quartz SAR OSL and post-IR infrared (IR) stimulated luminescence (post-IR IRSL; pIRIR290) measurements made on sand-sized quartz and K-feldspar extracts from the loess-palaeosol sequence at Niuyangzigou in northeastern China. The quartz OSL characteristics are satisfactory. Extensive pIRIR50,290 dose recovery tests were performed by adding doses on top of the natural dose. We found that dose recovery ratios improve significantly when the test dose ranges between ~15 and ~80% of the total dose, and good dose recovery (within ±5% of unity) can be obtained up to ∼800 Gy. Otherwise, the dose recovery ratio deviates from unity. The De values also depend on the test dose size and so we conclude that the effect of test dose size should be routinely considered in pIRIR dating. First IR stimulation plateau pIRIR290 results are compared with multiple elevated temperature -pIRIR (MET-pIRIR) data. It appears that the low temperature MET-pIRIR data are strongly affected by poor dose recovery, but this is not the case for the pIRIR290 results. Natural signal measurements at the highest (first IR) stimulation temperature on a sample expected to be in field saturation suggest that ~10% signal loss occurs in pIRIR signals. Long-term laboratory bleaching experiments (>80 days) show that a constant (or very difficult to bleach) residual pIRIR290 signal is reached after ~300 h, corresponding to a dose of 6.2±0.7 Gy. Quartz OSL and feldspar pIRIR50,290 ages are in good agreement at least back to c. 70 ka. Beyond this the quartz ages begin to underestimate but the feldspar ages are in agreement with the expected Last Interglacial age palaeosol.