Cortisol concentrations were measured in matched plasma and salivary samples from 8 healthy controls, 8 patients with Cushing's syndrome and 4 patients suspected of having spontaneous hypercortisolism. In healthy subjects, the circadian rhythm in salivary cortisol paralleled that in plasma. Absence of the diurnal rhythm in Cushing's syndrome was seen in saliva as well as in plasma. After ACTH stimulation, mean peak cortisol in saliva showed a 3-fold increase while in plasma there was a 2.5-fold increment above baseline. Cushing's syndrome, due to pituitary or adrenal adenoma was diagnosed equally well by measuring the cortisol response to cosyntropin in either plasma or saliva. Finally, the low- and high-dose dexamethasone suppression test was reflected equally well in both plasma and saliva. In patients suspected of having Cushing's syndrome dynamic tests can be performed in both plasma and saliva. However, in some samples, the salivary cortisol measurement appears advantageous over plasma cortisol determination.