We sense the temperature of our skin and surroundings using specific thermoreceptors, responding either to skin cooling or warming. Until a few years ago the molecular basis of thermal sensing was completely unknown. The discovery by Cesare and McNaughton in 1996 of an ionic current activated by noxious heat in cultured dorsal root ganglion (DRG) neurones revealed the first of a new family of temperature-gated ion channels, which we now know are all TRP channels. Here I will concentrate on one member of the family, the cold- and menthol-activated ion channel TRPM8, and its properties in native DRG neurones.Cold-sensitive DRG neurones are scarce – only 7 % of the total – and for this reason we pre-selected them using [Ca²⁺]_i imaging. In neurones showing a sudden sharp increase in [Ca²⁺]_i during cooling, we found an inward current activated by cold and sensitised by menthol, which looked like a “mirror image” of TRPV1’s activation by heat and sensitisation by capsaicin, and seemed an obvious candidate for a specific cold transduction mechanism (Reid & Flonta, 2001). The channel underlying this cold- and menthol-activated current turned out also to be a member of the TRP family, namely TRPM8 (McKemy et al., 2002; Peier et al., 2002).We had two major questions on discovering this current: firstly, to what extent could it account for cold transduction in intact thermoreceptors?; secondly, is it due to an ion channel that is directly activated by cold and menthol?To understand its function, we first looked at which of the known properties of intact cold receptors could be accounted for by this current. Intact cold receptors are sensitised by menthol and by low [Ca²⁺]_o; the effect of menthol is antagonised by high [Ca²⁺]_o and by warming; and intact cold receptors adapt to sustained cooling with a gradual reduction in their firing rate. The cold- and menthol-activated current shares all these properties, leading us to propose that it is probably the major transduction mechanism in innocuous thermoreceptors (Reid & Flonta, 2001).Excised patches from cold-sensitive DRG neurones contain a nonselective cation channel that is activated by cold and sensitised by menthol, accounting for the macroscopic current (Reid & Flonta, 2002). This confirmed that cold and menthol are acting on the same target, and that the channels involved are activated directly by cold and menthol and not by way of a soluble intracellular second messenger. However, channels in excised outside-out patches do not behave identically to those in intact neurones. Firstly, they are much less sensitive to cold; secondly, they no longer adapt to cold but instead show a sustained activity that depends only on temperature and does not change with time. We concluded that, whereas the temperature sensor and the menthol binding site are intrinsic to the channel (or an accessory subunit whose association with the channel resists patch excision), the adaptation mechanism is not contained within the channel itself but depends on something that is supplied by the cell (Reid & Flonta, 2002).The observations above indicate that the temperature threshold of cold receptors depends essentially on the modulation of TRPM8 and not only on its intrinsic thermal sensitivity. This matches well with our everyday sensory experience water at 20 °C feels very cold when we jump into it on a hot summer day, but very warm when we have just been throwing snowballs – and suggested that a shift in TRPM8’s cold activation threshold (as distinct from a simple decline in the current on sustained cooling) could account for cold receptor adaptation. During cold pulses of varying duration and intensity, we found that cooling indeed shifts the activation threshold of the cold- and menthol-activated current towards lower temperatures. Adaptation was strongly inhibited by removing extracellular Ca²⁺ or chelating intracellular Ca²⁺, and thermal threshold could be lowered without cooling simply by raising [Ca²⁺]_i (Reid et al., 2002). We concluded that the thermal threshold of cold receptors is modulated by Ca²⁺ entry through TRPM8, implying a feedback regulation of cold receptor sensitivity. Further work to be presented at this meeting shows that cold adaptation depends on membrane integrity but not on an intact cytoplasm. All procedures accord with the principles of UK legislation.