The relative effects of 3 stimuli in the release of vasopressin (VP) by the fetus were examined in 19 chronically instrumented fetal lambs, 118-135 days gestational age. The fetus was exposed to: 1) 30-min administration of 10% O2 to the pregnant ewe, 2) 20 min of partial occlusion of the umbilical cord, or 3) 2 min of complete occlusion of the umbilical cord. Twelve studies were conducted for each of these experimental protocols. The rises (mean +/- SE) in plasma VP in the 3 groups were 29.9 +/- 8.9, 48.9 +/- 11.5, and 157.8 +/- 12.5 pg/ml, respectively; the corresponding falls in PaO2 were 9.2 +/- 0.7, 7.7 +/- 0.9, and 12.7 +/- 1.2 mm Hg. pHa did not change in the group receiving 10% O2, fell by 0.11 +/- 0.02 and 0.14 +/- 0.01 after partial and complete occlusion of the umbilical cord. The rises in mean arterial pressure were 6 +/- 2.2, 10 +/- 1.9, and 23 +/- 3.1 mm Hg, respectively, at the end of the 3 procedures. The rise in plasma osmolality ranged from 2-8 mosmol/kg in all 3 groups. Linear regression analyses showed that log VP was negatively correlated with PaO2 (r = -0.827; P less than 0.01) and pHa (r = -0.706; P less than 0.01) and positively correlated to mean arterial pressure (r = 0.607; P = 0.01), but was not significantly correlated to plasma osmolality. Our present findings suggest that hypoxemia and acidemia are associated with and may both be potent stimuli for VP release in the hypoxic fetus. Increased release together with decreased rate of metabolism by the placenta offer an explanation for the very high VP levels found in the neonate after labor and delivery.