We evaluated the effects of prolonged spironolactone treatment on aldosterone secretion in patients with primary aldosteronism. The patients were hospitalized and underwent a furosemide test with or without dexamethasone, as well as an adrenocorticotrophin (ACTH) test. In untreated patients, neither plasma renin activity (PRA) nor plasma aldosterone showed a response in the furosemide test. In patients receiving spironolactone, furosemide increased significantly both the PRA and the plasma aldosterone concentration (from 2.6 +/- 0.8 to 7.0 +/- 2.0 micrograms.l-1.h-1 (p < 0.05) and from 345.6 +/- 55.8 to 492.7 +/- 76.8 ng/l (p < 0.05), mean +/- SEM, respectively). Dexamethasone administration had no effect on the results of the furosemide test (p > 0.1). However, dexamethasone tended to decrease the basal plasma aldosterone concentration in the untreated patients, but not in the patients receiving spironolactone. In the ACTH test, the plasma aldosterone concentration increased significantly in the untreated patients (from 549.0 +/- 69.8 to 1169.3 +/- 165.5 ng/l, p < 0.01), but there was no significant aldosterone response in the spironolactone-treated patients (from 885.5 +/- 204.9 to 1260.3 +/- 289.2 ng/l, p > 0.1). We conclude that aldosterone secretion is mainly dependent on ACTH in the untreated patients with primary aldosteronism and is more strongly regulated by the renin-angiotensin system during spironolactone treatment.