Alzheimer's disease is often characterized by an increase in plasma cortisol without clinical evidence of hypercorticism. Twenty-three consecutive patients with Alzheimer's disease and 23 age- and sex-matched healthy controls were studied by measuring plasma cortisol and dehydroepiandrosterone sulfate (DHEAS) (by enzyme immunoassay), the number of type I and type II corticosteroid receptors in mononuclear leukocytes (by radio-receptor-assay), and the lymphocyte subpopulations (by cytofluorimetry). Results are expressed in terms of median and range. In Alzheimer's disease, plasma cortisol was higher than in controls (median 0.74, range 0.47-1.21 vs 0.47, 0.36-0.77 mmol/L; p < 0.001). Plasma DHEAS, the DHEAS/cortisol ratio, and the number of type II corticosteroid receptors were significantly lower in AD than in controls (DHEAS: median 1.81, range 0.21-3.69 vs 3.51, 1.35-9.07 micromol/L; DHEAS/ cortisol: 2.04, range 0.3-5.8 vs 6.8, range 2.7-24 and type II receptors: 1219, 1000-2700 vs 1950, 1035- 2750 receptors per cell; p < 0.001). No correlation was found between the hormonal parameters, age, and mini-mental test score. These data support the hypothesis of a dysregulation of the adrenal pituitary axis in Alzheimer's disease, which is probably the consequence of damage to target tissues by corticosteroids.