The aim of the present study was to investigate the hormonal control of water-balance in children with diabetes insipidus and to assess safety and efficacy of long-term treatment with oral dDAVP. Plasma atrial natriuretic peptide, plasma renin activity, aldosterone, plasma and urinary cyclic 3'5'-guanosine monophosphate and urinary prostaglandin E2 were measured in eight patients (aged 3-21 y) with central diabetes insipidus. At baseline, 12 h after the last dDAVP dose, patients had hypotonic polyuria but normal plasma sodium concentrations and plasma osmolality relative to a control group. The mean plasma atrial natriuretic peptide concentration in patients (26.2 +/- 2.6 pg/ml) tended to be lower than in controls (36.5 +/- 8.2 pg/ml, mean +/- SEM), although the difference was not significant. Plasma cyclic 3'5' guanosine monophosphate was higher in controls (6.0 +/- 0.6 pmol/ml, mean +/- SEM) than in patients (3.8 +/- 0.3 pmol/ml). Aldosterone, plasma renin activity, urinary cyclic guanosine monophosphate and urinary prostaglandin E2 were similar in the two groups. During 3 h following dDAVP administration, atrial natriuretic peptide levels did not change in patients but decreased significantly in controls to 23.0 +/- 4.0 pg/ml. No adverse reactions, or circulating antibodies against dDAVP, were observed after 3.5 years of oral dDAVP treatment. The average oral dDAVP dosage was similar after 1 and 3.5 years of treatment (906 +/- 406 micrograms/24 h, mean +/- SD). Water-balance is not detectably different from normal in correctly treated diabetes insipidus patients in terms of plasma atrial natriuretic peptide, plasma renin activity and aldosterone levels. Long-term oral dDAVP treatment is safe and efficacious.