Thermally transferred optically stimulated luminescence (TT-OSL) dating protocols have been suggested as a means of extending the age range of luminescence dating. Several studies demonstrate that TT-OSL signals increase with large radiation doses (>2000 Gy) and yet, few studies report older TT-OSL ages (>400 Gy) in agreement with independent absolute age control. In one such study, agreement with independent chronology was only achieved for the old samples by implementing a pulsed-irradiation procedure. Pulsed-irradiation is suggested to compensate for dose rate dependent competition effects by dividing the laboratory irradiation into discrete irradiation steps interspersed with heat treatments. However, every inter-step heat treatment has the potential to anneal part of the TT-OSL dating signal. This study compares constant- and pulsed-irradiation TT-OSL protocols and investigates the degree of partial thermal annealing. The results suggest that almost all of the difference in outcome between constant- and pulsed-irradiation protocols can be explained by partial annealing of the TT-OSL signal rather than by competition effects. Partial annealing distorts the laboratory dose response curve but has no impact on the natural signal, resulting in unreliable equivalent dose estimates. This means that pulsed-irradiation procedures may not be viable for TT-OSL dating measurements. Future studies implementing pulsed-irradiation procedures should carefully consider the extent to which inter-step thermal treatments partially anneal the signal.