The human heart secretes ANP, mainly or exclusively as the 1-28-amino-acid alpha-hANP. Secretion is increased when there is hypervolemia of the central circulation, either acute or chronic, the stimulus being, it is presumed, atrial stretch. The clearance rate of alpha-hANP has been documented in healthy volunteers but not in patients with clinical disorders. Injection or infusion of atrial peptides into normal humans has clear-cut hemodynamic, renal, and hormone effects. However, the doses used have been high and the results do not allow extrapolation to the realms of physiology. Patients with essential hypertension appear to have exaggerated renal responses to administered alpha-hANP, although the number of subjects studied is small and matching with normotensive controls was imperfect. By contrast, cardiac failure is characterized by impaired renal responses to atrial peptides. The place of atrial peptides in human physiology and pathophysiology is not clear and will require very low-dose infusion studies under exacting experimental conditions or the development of a specific inhibitor of circulating ANP. In theory, atrial peptides might find a place in therapeutics, most likely in cardiac failure or renal failure, but also essential hypertension if an orally active agonist can be developed.