It has been established that in vivo administration of bacterial lipopolysaccharide (LPS) enhances hypothalamo-pituitary-adrenal (HPA) axis function by a mechanism involving endotoxin-stimulated cytokine release. Since under chronic LPS treatment a tolerance of the HPA axis response takes place, the aim of the present study was to determine whether mice submitted to repeated LPS administration could present an impairment in the HPA response to insulin (INS) administration, a pure neuroendocrine challenge. For this purpose, adult female BALB/c mice were injected with 200 microliters i.p. of sterile saline solution (VEH) containing 25 micrograms of LPS in a single or repeated (at 24-hour intervals, during 5 consecutive days) fashion. Animals were then killed at either 45 min after INS (0.3 IU/mouse, i.p.) or 2 h after LPS (25 micrograms/mouse) administration on experimental day 1 (D1; without any previous LPS injection), 3 (D3; mice having received 2 previous LPS injections) or 5 (D5; mice having received 4 previous LPS administrations). Control groups were injected a similar volume of VEH alone on experimental day 1 (D1; without any previous LPS injection), 3 (D3; mice having received 2 previous LPS injections) or 5 (D5; mice having received 4 previous LPS administrations); in each group, mice were killed at either 45 min or 2 h after VEH injection. Immediately after decapitation, trunk blood was collected. Plasma tumor necrosis factor alpha (TNF), ACTH and corticosterone (B) levels were determined by specific assays. Plasma TNF, ACTH and B levels were significantly increased 2 h after the first LPS treatment (D1). Although no significant increase in plasma TNF concentration was found 2 h after the third LPS injection (D3), the corticotrope response was still significant and induced a full effect on adrenal B output. Two hours after the fifth LPS administration (D5) TNF output was minimal and the HPA axis response was significantly diminished. Finally, the pattern of the HPA axis response to INS-induced hypoglycemia was similar to that elicited after LPS challenge although somewhat delayed. Our results indicate that: (1) TNF seems to play an important role in stimulating HPA axis function after single but not after repeated endotoxin administration; and (2) an impairment in the HPA axis response to both immuneneuroendocrine (LPS) and neuroendocrine (INS) stimuli takes place after repeated LPS administration. This study further suggests that the tolerance of the HPA axis response under recurrent endotoxemia could be, at least partially, due to an impairment in both immune (TNF output) and neuroendocrine functions.