The temperatures of maximum thermoluminescence, as recorded by peaks on glow curves, do not mark valid points for use in geothermometry. These peak temperatures are determined by the heating rate and characteristics of the thermoluminescence apparatus. Consequently, the maximum and minimum limits of temperature estimated for a thermal aureole are not defined by the peak temperatures. Although the amount of thermoluminescence measured in a sample reflects an equilibrium between radioactive bombardment and concomitant electron leakage, the complete drainage of an electron trap seems to take place only above a characteristic threshold level. Each electron trap can be drained completely at or above the temperature at which thermoluminescence begins. For example, the thermoluminescence in a limestone sample with "peak temperatures" of 245 degrees C. and 320 degrees C. can be drained at 150 degrees -155 degrees C. and 205-210 degrees C., respectively. These temperatures of initial luminescence should be used in geothermometry, instead of the peak temperatures. The application of thermoluminescence geothermometry to heat aureoles near hydrothermal ores and intrusives is not encouraging. Consistent results are obtained only where the rocks are abnormally homogeneous. Moreover, even homogeneous wall rocks near ore deposits do not have suppressed thermoluminescence. Pre-Pleistocene heating effects are masked by subsequent radioactive bombardment and the re-establishment of equilibrium between the rates of trap filling and electron leakage. A higher saturation level is likely to be developed in the thermally metamorphosed zone. As a result, the application of thermoluminescence geothermometry near most ore deposits is questionable. Possibly future discoveries concerning the theory of thermoluminescence or new refinements in technique will result in the use of this phenomenon as a reliable geothermometer.