The multiple effects of vagotomy on the thermoregulatory response to systemic inflammation are reviewed (primarily, for the model of intravenous lipopolysaccharide administration in the rat). The following conclusions are drawn. (1) Vagotomy-associated thermoeffector insufficiency is likely to account for the attenuation of the fever response observed in some--but not all--studies; such an insufficiency is, however, preventable by postoperative care, including the use of a liquid diet. (2) The febrile response to low doses of lipopolysaccharide (monophasic fever) is mediated by the hepatic (but not gastric or celiac) vagal fibers, presumably afferent; the same fibers are likely to be involved in the development of tolerance to low doses of circulating endotoxins. (3) Phase 1 of the polyphasic febrile response to moderate doses of lipopolysaccharide involves capsaicin-sensitive afferents (either nonvagal only or both nonvagal and vagal), does not involve cholecystokinin A-receptors, and may involve peripheral prostaglandins. (4) Febrile phase 2 does not require the integrity of abdominal nerve fibers, either vagal or nonvagal, at least in the rat. (5) Phase 3 of the febrile response to intravenous lipopolysaccharide (and perhaps the response to intraperitoneal lipopolysaccharide) involves capsaicin-insensitive vagal fibers, presumably efferent; the involvement of these fibers in febrigenic mechanisms is strongly modulated by an unknown factor. (6) A hepatoceliac vagal, presumably efferent, mechanism ('an anti-inflammatory pathway') counteracts the development of lipopolysaccharide-induced hypothermia and shock.