Human epidemiological data show a strong association between low birth weight and hypertension in adulthood, an effect that has been ascribed to 'fetal programming'. In rats fetoplacental exposure to maternally administered dexamethasone throughout gestation reduces birth weight and produces hypertensive adult offspring, though the mechanism is unclear. Pre- and postnatal stress programmes hypothalamic-pituitary-adrenal (HPA) axis responses throughout the lifespan, an effect thought to be mediated via permanent effects on glucocorticoid receptor (GR) and/or mineralocorticoid receptor (MR) gene expression in the hippocampus. Corticosteroids also have specific central effects on blood pressure control mediated by GR and MR. This study investigated corticosterone (CORT) responses to restraint stress and GR and MR gene expression in areas of the brain postulated to mediate the central effects of corticosteroids on (i) HPA axis suppression (hippocampus), and (ii) blood pressure (organ vasculosum of the lamina terminalis (OVLT), sub-commissural organ, area postrema and nucleus tractus solitarius). Pregnant Wistar rats received dexamethasone (100 micrograms/kg.day-1) or vehicle on days 15-20 of gestation. This reduced birth weight by 11%. When the offspring were 16 weeks old, blood pressure was recorded directly and plasma CORT measured basally (AM) and after 30 min restraint. GR and MR mRNA expression were determined by in situ hybridization. Blood pressure was significantly elevated in the adult offspring of dexamethasone-treated pregnancies (dexamethasone 144 +/- 2/125 +/- 2 mm Hg vs. control 133 +/- 2.7/112 +/- 2.8 mm Hg; both p < 0.01). Offspring of dexamethasone-treated pregnancies had increased basal plasma CORT (155 +/- 29 nmol/l) compared to offspring of controls (79 +/-15 nmol/l, p < 0.05), but the CORT response to stress was similar. Hippocampal neuronal GR mRNA expression was significantly lower in the offspring of dexamethasone-treated pregnancies (dentate gyrus 20% lower, CA1 15% lower; p < 0.01). Similarly, hippocampal MR gene expression was decreased in CA1 and CA2 by 24 and 25%, respectively (p < 0.05). No differences in GR or MR mRNA expression were found in the OVLT, subcommissural organ, area postrema or nucleus tractus solitarius. These findings suggest that glucocorticoid excess in the last trimester of rat pregnancy (i) is sufficient to programme offspring hypertension; (ii) also increases basal plasma CORT levels, and (iii) permanently attenuates GR and MR mRNA expression in specific hippocampal subfields. Thus, if translated into protein, may reduce sensitivity to glucocorticoid feedback and thus contribute to the CORT excess. However, hypertension in this model is unlikely to be mediated by similar changes in GR or MR gene expression in the examined areas of the brain putatively involved in the more direct central regulation of blood pressure.