The innate immune system serves as the first line of host defense against the deleterious effects of invading infectious pathogens. Fever is the hallmark among the defense mechanisms evoked by the entry into the body of such pathogens. The conventional view of the steps that lead to fever production is that they begin with the biosynthesis of pyrogenic cytokines by mononuclear phagocytes stimulated by the pathogens, their release into the circulation and transport to the thermoregulatory center in the preoptic area (POA) of the anterior hypothalamus, and their induction there of cyclooxygenase (COX)-2-dependent prostaglandin (PG)E(2), the putative final mediator of the febrile response. But data accumulated over the past 5 years have gradually challenged this classical concept, due mostly to the temporal incompatibility of the newer findings with this concatenation of events. Thus, the former studies generally overlooked that the production of cytokines and the transduction of their pyrogenic signals into fever-mediating PGE(2) proceed at relatively slow rates, significantly slower certainly than the onset latency of fever produced by the i.v. injection of bacterial endotoxic lipopolysaccharides (LPS). Here, we review the conflicts between the earlier and the more recent findings and summarize new data that reconcile many of the contradictions. A unified model based on these data explicating the generation and maintenance of the febrile response is presented. It postulates that the steps in the production of LPS fever occur in the following sequence: the immediate activation by LPS of the complement (C) cascade, the stimulation by the anaphylatoxic C component C5a of Kupffer cells, their consequent, virtually instantaneous release of PGE(2), its excitation of hepatic vagal afferents, their transmission of the induced signals to the POA via the ventral noradrenergic bundle, and the activation by the thus, locally released norepinephrine (NE) of neural alpha(1)- and glial alpha(2)-adrenoceptors. The activation of the first causes an immediate, PGE(2)-independent rise in core temperature (T(c)) [the early phase of fever; an antioxidant-sensitive PGE(2) rise, however, accompanies this first phase], and of the second a delayed, PGE(2)-dependent T(c) rise [the late phase of fever]. Meanwhile-generated pyrogenic cytokines and their consequent upregulation of blood-brain barrier cells COX-2 also contribute to the latter rise. The consecutive steps that initiate the febrile response to LPS would now appear, therefore, to occur in an order different than conceived originally.