Prenatal stress or glucocorticoid administration has persisting "programming" effects on offspring in rodents and other model species. Multiple doses of glucocorticoids are in widespread use in obstetric practice. To examine the clinical relevance of glucocorticoid programming, we gave 50, 120, or 200 microg/kg/d of dexamethasone (dex50, dex120, or dex200) orally from mid-term to a singleton-bearing nonhuman primate, Chlorocebus aethiops (African vervet). Dexamethasone dose-dependently reduced maternal cortisol levels without effecting maternal blood pressure, glucose, electrolytes, or weight gain. Birth weight was unaffected by any dexamethasone dose, although postnatal growth was attenuated after dex120 and dex200. At 8 months of age, dex120 and dex200 offspring showed impaired glucose tolerance and hyperinsulinemia, with reduced (approximately 25%) pancreatic beta cell number at 12 months. Dex120 and dex200 offspring had increased systolic and diastolic blood pressures at 12 months. Mild stress produced an exaggerated cortisol response in dex200 offspring, implying hypothalamic-pituitary-adrenal axis programming. The data are compatible with the extrapolation of the glucocorticoid programming hypothesis to primates and indicate that repeated glucocorticoid therapy and perhaps chronic stress in humans may have long-term effects.