The injection of a melanocortin peptide or of melanocortin peptide analogues into the cerebrospinal fluid or into the ventromedial hypothalamus in nanomolar or subnanomolar doses induces a long-lasting inhibition of food intake. The effect keeps significant for up to 9 h and has been observed in all animal species so far tested, the most susceptible being the rabbit. The anorectic effect of these peptides is a primary one, not secondary to the shift towards other components of the complex melanocortin-induced behavioral syndrome, in particular grooming. The site of action is in the brain, and the effect is not adrenal-mediated because it is fully exhibited also by adrenalectomized animals. It is a very strong effect, because the degree of feeding inhibition is not reduced in conditions of hunger, either induced by 24 h starvation, or by insulin-induced hypoglycemia, or by stimulation of gamma-aminobutyric acid (GABA), noradrenergic or opioid systems. The microstructural analysis of feeding behavior suggests that melanocortins act as satiety-inducing agents, because they do not significantly modify the latencies to start eating, but shorten the latencies to stop eating. The mechanism of action involves the activation of melanocortin MC(4) receptors, because selective melanocortin MC(4) receptor antagonists inhibit the anorectic effect of melanocortins, while inducing per se a strong stimulation of food intake and a significant increase in body weight. Melanocortins seem to play an important role in stress-induced anorexia, because such condition, in rats, is significantly attenuated by the blockage of melanocortin MC(4) receptors; such a role is not secondary to an increased release of corticotropin-releasing factor (CRF), because, on the other hand, the CRF-induced anorexia is not affected at all by the blockage of melanocortin MC(4) receptors. The physiological meaning of the feeding inhibitory effect of melanocortins, and, by consequence, the physiological role of melanocortins in the complex machinery responsible for body weight homeostasis, is testified by the hyperphagia/obesity syndromes caused by mutations in the pro-opiomelanocortin (POMC) gene, or in the melanocortin MC(4) receptor gene, or in the agouti locus. Finally, recent evidences suggest that melanocortins could be involved in mediating the effects of leptin, and in controlling the expression of neuropeptide Y (NPY).