Metabolic clearance rates (MCR) of arginine vasopressin (AVP) were measured serially in five women starting before conception, during gestational weeks 7-8 (early), 22-24 (middle), and 36-38 (late pregnancy), and again 10-12 wk postpartum. Hormonal disposal rates were determined after water loading to suppress endogenous AVP release using a constant infusion method designed to achieve three different steady-state concentrations of plasma AVP (PAVP) on each test occasion. Dose schedules were altered in mid- and late pregnancy to obtain comparable AVP levels at each stage of the protocol. Prehydration decreased plasma osmolality sufficiently to suppress AVP release, as circulating AVP-neurophysin measured serially in three of the women was undetectable. The MCR of AVP was similar before conception (0.75 +/- 0.31, 0.79 +/- 0.34, and 0.76 +/- 0.28 liters/min at PAVP of 2.6 +/- 1.9, 4.7 +/- 2.4, and 8.3 +/- 3.9 pg/ml), in early pregnancy (0.89 +/- 0.34, 0.97 +/- 0.04, and 0.95 +/- 0.40 liters/min at PAVP of 2.2 +/- 2.1, 3.9 +/- 3.2, and 7.9 +/- 3.4 pg/ml), and postpartum (0.70 +/- 0.21, 0.69 +/- 0.24, and 0.75 +/- 0.20 liters/min at PAVP 3.5 +/- 1.8, 5.1 +/- 3.7, and 9.1 +/- 4.2 pg/ml). Values at mid-pregnancy (2.8 +/- 1.3, 3.0 +/- 1.2, and 2.7 +/- 1.2 liters/min at PAVP 2.3 +/- 2.2, 4.0 +/- 3.6, and 7.7 +/- 3.9 pg/ml) and late pregnancy (3.2 +/- 1.4, 3.3 +/- 1.4, and 2.9 +/- 1.2 liters/min at PAVP 1.9 +/- 2.0, 3.8 +/- 2.6, and 7.4 +/- 4.1 pg/ml) increased 3-4-fold (all P less than 0.01). Plasma vasopressinase, undetectable at 7-8 gestational wk, increased markedly by mid- and slightly more by late gestation. Finally, relationships between PAVP and urine osmolality were similar before, during, and after pregnancy. We conclude that marked increments in the MCR of AVP occur between gestational weeks 7 and 8 and mid-pregnancy, which parallel the period of greatest rise in both trophoblastic mass and plasma vasopressinase. There was no evidence of a renal resistance to AVP during gestation.