Maternal and umbilical cord serum sodium and osmolality were studied prospectively in 140 deliveries to investigate whether transplacental hyponatraemia, seen following oxytocin infusion during labour, was due to the antidiuretic effect of oxytocin or was secondary to the infusion of aqueous glucose used as a vehicle for oxytocin, or both. Forty-five women received oxytocin in aqueous glucose for induction or augmentation of labour (oxytocin group), 43 received aqueous glucose infusion alone (glucose group) and 52 did not receive any intravenous infusions (control group). Mean cord sodium levels were significantly lower in the oxytocin (131.4, SD 3.6 mmol/l) and glucose groups (132.5, SD 3.2 mmol/l) than in the control group (135.0, SD 3.0 mmol/l). Hyponatraemia (Na less than 130 mmol) was seen in 47% and 30% of the infants in the oxytocin and glucose groups respectively, in contrast to only 5.8% of the infants in the control group. Significant negative linear correlations were seen between serum sodium and the dose of oxytocin (P less than 0.01) and log of the volume of glucose solution infused (P less than 0.001). The hyponatraemic newborn infants had a significantly higher incidence of transient neonatal tachypnea (7/37, 19%) than the normonatraemic infants (2%). Our results strongly suggest that infusion of oxytocin and glucose both cause maternal and transplacental hyponatraemia, even in recommended doses. This should be taken in account while planning a safe dose of oxytocin and glucose for infusion during labour.