In track structure theory (TST), the calculation of the relative thermoluminescence (TL) efficiency of heavy charged particles (HCPs) is carried out by integration of radial dose profiles weighted with the TL efficiency measured by a low ionization density dose response. The reliability of these calculations is reviewed herein with special emphasis on the composite glow peak 5 system in LiF : Mg, Ti (TLD-100). Possible weaknesses in the basic assumptions and in the details of the experimental and calculational methods are discussed. The influence of HCP-enhanced production of atomic displacements/defects on the TL efficiency is investigated via an experiment sensitive to F-centre concentrations which may influence the luminescence mechanism/efficiency. In this experiment, the sample is irradiated by 5.3 MeV alpha particles to a low fluence of ∼108 cm−2 in order to minimize track overlap and then read out to a maximum temperature of 350 °C. A second readout is carried out following irradiation by 250 nm photons corresponding to the 5 eV optical excitation energy of the F centre. The intensity of glow peak 5 in the second readout relative to the first readout is calculated theoretically using the first-readout and second-readout dose responses of peak 5 measured with both 100 keV x-rays and 90Sr/90Y beta rays and found to lie outside the predictive capability of TST.