Prolonged stress exposure threatens early pregnancy maintenance and can lead to spontaneous abortion (1). In the abortion-prone DBA/2J-mated CBA/2J female mice, 24 h noise stress on day 5.5 of pregnancy increases the number of resorptions of implanted embryos by elevating pregnancy-threatening cytokines (2). This is caused by a decrease in progesterone secretion and is prevented by replacement with a progesterone analogue (dydrogesterone) at the time of stress (3). We have now investigated how the hypothalamo-pituitary-adrenal (HPA) axis responds in early pregnant c57/Bl6 mice by analysing immediate early gene expression in the paraventricular nucleus (PVN) and peripheral secretory responses. Since the stress-induced abortion in the above model is mediated by cytokines, we used an immune stress in early pregnant mice by administering the endotoxin, lipopolysaccharide (LPS). Virgin and 5.5 day and 10.5 day pregnant mice were injected i.p. with LPS (12.5 μg per mouse, n=9,6,5, respectively) or vehicle (100 μl isotonic saline, n=8,7,4, respectively) and decapitated 90 min later when blood samples and brains were collected. Brains were analysed for nur77 (immediate early gene associated with HPA axis activation) expression by in situ hybridisation and plasma was separated and analysed for adrenocorticotrophin (ACTH) and progesterone concentration by RIA; data are mean±S.E.M. LPS increased nur77 mRNA expression in the PVN in terms of grain area per cell in virgin mice compared to vehicle (13.6±1.3 vs. 7.6±1.0 μm²), but expression was significantly lower in LPS-treated pregnant groups (9.3±1.4 and 8.6±0.7 μm²) vs. virgins (p<0.05 across group and treatment, 2-way ANOVA). ACTH concentration was greater in all groups after LPS compared to vehicle and was significantly greater on day 10.5 of pregnancy compared to virgins (781.3±47.7 vs. 636.7±38.0 pg/ml; p<0.05 1-way ANOVA). On both days of pregnancy control mice had significantly greater progesterone concentration than control virgin mice (pregnancy: 35.9±7.8 and 45.5±10.7 vs. 4.8±0.8 pg/ml). In the virgin mice LPS increased progesterone secretion (to 18.3±5.9 pg/ml). However, in contrast, in the pregnant groups progesterone concentration was less after LPS compared to vehicle (day 5.5, 20.3±2.0 and day 10.5, 19.2±2.4 pg/ml; p<0.05 across group and treatment, 2-way ANOVA). So, in early pregnancy there was an attenuated PVN response to LPS but an acute exaggerated ACTH secretory response, indicating dissociation between central and peripheral control mechanisms. On the other hand, there was a stress-induced inhibition of sex steroid secretion. Thus, immune signals rapidly inhibit progesterone secretion in early pregnancy, an effect which may be prolonged and contribute to stress-induced abortion. While central responses in the PVN were reduced at this time, their role, if any, in the regulation of the hypothalamo-pituitary-gonadal axis and corpus luteum progesterone secretion remains to be investigated.