An unexplained hallmark of prolonged critical illness is the fact that food does not prevent or reverse protein wasting, while fat is paradoxically accrued. This`wasting syndrome' often persists after the underlying disease has been resolved and thus perpetuates intensive care dependency. Although the crucial role of an intact hypothalamus±pituitary axis for homeostasis during stress is well recognized, the differences between the neuroendocrine changes observed in acute and prolonged critical illness were only recently described. Novel insights in this area are reviewed here.The initial endocrine stress response consists primarily of a peripheral inactivation of anabolic pathways while pituitary activity is essentially ampli®ed or maintained. These responses presumably provide the metabolic substrates and host defense required for survival and to delay anabolism, and thus should be considered as adaptive and bene®cial. Persistence of this acute stress response throughout the course of critical illness was hitherto assumed. This assumption has now been invalidated, since a uniformly reduced pulsatile secretion of ACTH, TSH, LH, prolactin (PRL) and GH has been observed in protracted critical illness, causing diminished stimulation of several target organs. Impaired pulsatile secretion of anterior pituitary hormones in the chronic phase of critical illness seems to have a hypothalamic rather than a pituitary origin, as administration of relevant releasing factors evoked immediate and pronounced pituitary hormone release. A reduced availability of TRH, one of the endogenous ligands of the GH-releasing peptide (GHRP) receptor (such as the recently discovered ghrelin) and, in very long-stay critically ill men, also of GHRH, appear to be involved. This hypothesis was further explored by investigating the effects of continuous i.v. infusion of GHRH, GHRP, TRH and their combinations for several days. Pulsatile secretion of GH, TSH and PRL was re-ampli®ed by relevant combinations of releasing factors which also substantially increased circulating levels of IGF-I, GH-dependent binding proteins, thyroxine and tri-iodothyronine (T3) while avoiding a rise in reverse T3. Active feedback-inhibition loops prevented overstimulation of target organs and metabolic improvement was noted with the combined infusion of GHRP and TRH. Whether this novel endocrine strategy will also enhance clinical recovery from critical illness remains to be explored.