A pre-requisite to understanding the physiological mechanisms of action of melatonin is the identification of the target sites where the hormone acts. The radioligand 2-[125I]iodo-melatonin has been used extensively to localize binding sites in both the brain and peripheral tissues. In general these binding sites have been found to be high affinity, with Kd in the low picomolar range, and selective for structural analogues of melatonin. Also the affinity of these sites can generally be modulated by guanine nucleotides, consistent with the notion that they are putative G-protein coupled receptors. However, only a few studies have demonstrated that these putative receptors mediate biochemical and cellular responses. In the pars tuberalis (PT) and pars distalis (PD) of the pituitary, the amphibian melanophore and vertebrate retina, evidence indicates that melatonin acts to inhibit intracellular cyclic AMP through a G-protein coupled mechanism, demonstrating that this is a common signal transduction pathway for many melatonin receptors. However in the pars distalis the inhibition of calcium influx and membrane potential are also important mediators of melatonin effects. How many different forms or states of the melatonin receptor exist is unknown, but clearly the identification of the structure of the melatonin receptor(s) and its ability to interact with different G-proteins and signal transduction pathways are quintessential to our understanding of the physiological mechanisms of action of melatonin. In parallel the recent development of new melatonin analogues will greatly aid our understanding of the pharmacology of the melatonin receptor both in terms of the development of potent melatonin receptor antagonists and for the definition of receptor sub-types. The wide species and phylogenic diversity of melatonin binding sites in the brain has probably generated more questions than answers. Nevertheless the localization of melatonin receptors to the suprachiasmatic nucleus of the hypothalamus is at least consistent with circadian effects within the foetus and the adult. In contrast the PT of the pituitary presents an enigma in relation to the seasonal effects of melatonin. A model of how melatonin might mediate the timing of the circannual events through the PT is proposed.