Aims: Stress mechanisms paradoxically contribute to allergic episodes in humans and mice. Glucocorticoids (GC) and interleukin (IL)-5 synergically upregulatemurine bone-marroweosinophil production. Herewe explored the role of endogenous GC in allergen-stimulated bone-marrow eosinophil production in ovalbumin-sensitized/ challenged mice. Main methods: In BALB/c or C57BL/6 mice, sensitized and intranasally challenged with ovalbumin, we monitored eosinophil numbers in freshly harvested or cultured bone-marrow, and plasmacorticosterone levels.Metyrapone (MET) was used to inhibit GC synthesis, and RU486 to block GC actions. In sensitized mice challenged intraperitoneally, we examined the relationship between eosinophilia of bone-marrow and peritoneal cavity, in the absence or presence of RU486. In experiments involving in vivo neutralization of tumor necrosis factor-α (TNF) by specific antibodies, or using mice which lack functional type I TNF receptors (TNFRI), we evaluated the relationship between TNF blockade, corticosterone levels, RU486 or MET treatment and challenge-induced bone-marrow eosinophilia. Key findings: RU486 orMET pretreatments abolished challenge-induced increases in eosinophil numbers in bonemarrow( in vivo and ex vivo), and in the peritoneal cavity.MET, but not RU486, prevented the challenge-induced increase in corticosterone levels. Challenge-induced bone-marrow eosinophilia and corticosterone surge were abolished in TNFRI-deficient mice. Anti-TNF-treatment very effectively prevented challenge-induced bone-marrow eosinophilia, in the absence of RU486 or MET, but had no independent effect in the presence of either drug. Significance: Endogenous GC was essential for allergen challenge-induced increases in eosinophil numbers inside bone-marrow. This effect required TNF and TNFRI, which suggests an immunoendocrine mechanism.