To find out whether plasma vasopressin (PAVP) response to a water load during pregnancy is inappropriately high, as had been speculated, we measured PAVP by radioimmunoassay in 30 women at the time of delivery. Ten women had received infusion of aqueous glucose solution during labor for hydration (GW group); another ten received infusion of glucose solution as a vehicle for oxytocin (OT group), and ten women did not receive any intrapartum intravenous fluid therapy (controls). Serum sodium and osmolality were also determined in all the subjects. PAVP levels were significantly lower in GW (0.70 +/- 0.4 pg/ml) and OT groups (0.7 +/- 0.6 pg/ml) (P less than 0.05). Significant negative correlation was seen between the amount of glucose solution infused and levels of PAVP (r = -0.66; P less than 0.01), while a significant positive correlation was seen between PAVP and serum sodium (r = 0.61; P less than 0.01). These findings suggest that during labor, the physiological relationship between serum osmolality and PAVP is intact, and that infusion of a water load in the form of aqueous glucose solution is attended by an expected lowering of PAVP. We infer that inappropriate ADH response is not the cause of water retention and hyponatremia often seen in women receiving aqueous glucose solution during labor.