The finding that urine cortisol excretion was increased in patients with hypokalaemic hypertension induced by licorice addiction led to this study on the effect of licorice in normal subjects. Thirteen normal volunteers ate either 100 or 200 g licorice for 1-4 weeks and assessment of pituitary-adrenal function was made before, during, and 1 week after cessation of licorice ingestion. Urine cortisol excretion more than doubled in 10 of 13 subjects (mean, 33.8 +/- 15.6 SD before and 83.3 +/- 56 SD micrograms/24 h at 1 week after commencing licorice) and excretion rates similar to those observed in Cushing's syndrome were seen in 7 subjects (range, 91-226, compared to normal range of 11-82 micrograms/24 h). Urine cortisol excretion remained significantly elevated (P less than 0.01) above control levels for at least 1 week after licorice was withdrawn. Despite these increases, urinary steroid metabolite (tetrahydrocortisol, tetrahydrocortisone, tetrahydro-11-deoxycortisol, 17-ketogenicsteroids, and 17-ketosteroids) excretion was not affected, plasma cortisol and ACTH values were unchanged, and normal 0800-1600-h diurnal variation of plasma cortisol was maintained. The direct intraadrenal infusion of the active mineralocorticoid component of licorice, glycyrrhetinic acid, in two sheep with autotransplanted adrenal glands failed to stimulate cortisol secretion acutely. It is concluded from these studies that the licorice-induced changes in cortisol excretion are not a result of adrenocoritcal stimulation but more likely represent a change in the renal handling of cortisol.