The effectiveness of therapy with carbamazepine and clofibrate (oral therapy), intramuscular pitressin-in-oil, and intranasal 1-deamino-8-D-arginine vasopressin has been compared in 15 children with partial or complete central diabetes insipidus. Mean daily urine volume without therapy was 5.4 l and dropped to 1.1 and 1.6 l/day while receiving pitressin and DDAVP, respectively. Oral agents decreased the daily urine volume to 2.2 l in patients with partial DI, with good symptomatic control except for some nocturia. These agents had no effect in patients with complete DI and did not alter pitressin requirements. Duration of pitressin action was 24 to 36 hours with a significant incidence of hyponatremia. The duration of DDAVP effect was 8 to 20 hours, varying in individual patients. Children with partial DI required smaller doses of DDAVP and the duration of action was longer than in those with complete DI. Control of serum electrolytes was excellent using two doses per day and nocturia was eliminated. All patients who had received pitressin had growth hormone antibodies, but continued to grow normally unless there was pre-existing growth hormone deficiency. These antibodies gradually disappeared after approximately one year of therapy with oral agents or DDAVP. DDAVP did not alter growth hormone, cortisol, or prolactin levels during sleep. DDAVP is the antidiuretic therapy of choice in children with either complete or partial DI; to date, no side effects have been demonstrated.