Abstract The TL sensitivity of meteorites reflects the abundance and nature of the feldspar. Thus intense shock, which destroys feldspar, causes the TL sensitivity to decrease by 1–2 orders of magnitude, while metamorphism, which generates feldspar through the devitrification of primary igneous glass, causes TL sensitivity to increase by a factor of ∼ 105. TL-metamorphism relationship is particularly strong for the lowest levels of metamorphism and has led to a new means of classification for little-metamorphosed meteorites. The TL emission characteristics of feldspar (peak temperature and peak width) are also related to the thermal history. Induced TL properties have thus found useful application to the paleothermometry of primitive meteorites, to the thermal histories of chondrules, to the ejection mechanism of martian meteorites and to arguments about different extraterrestrial sources for meteorites currently falling world-wide and those which fell 105–106 yr ago in the Antarctic. The level of natural TL in meteorites is determined by competition between build-up, due to exposure to cosmic radiation, and thermal decay. Terrestrial conditions should result in lower levels of natural TL compared to those expected in space, and, in principle, natural TL levels should be related to the period of time on earth (“terrestrial age”). Antarctic meteorites tend to have lower natural TL than non-Antarctic meteorites whose fall was observed, presumably because of the large terrestrial ages of the latter (105–106 yr). A few meteorites have especially low natural TL suggesting that they have recently been reheated. These may have come to earth on orbits with particularly small perihelia (≤ 0.8 astronomical units), and 26 meteorites with calculated orbits show a correlation between natural TL and perihelion distance. Dose-rate variations due to shielding, heating during atmospheric passage and anomalous fading also cause variations in natural TL levels, but the effects are either relatively small, occur infrequently or can be experimentally identified.