In 50 healthy mothers scheduled for elective Caesarean section, anaesthesia was induced with propanidid (7 mg/kg body weight). Thereafter, ventilation was controlled with nitrous oxide, oxygen and muscle relaxants. A further dose of propanidid (1 mg/kg body weight) was administered 3 minutes after the initial injection of this drug, as a means of preventing maternal awareness during equilibration with the anaesthetic gas mixture. The acid-base status of the mothers before the induction of anaesthesia, and at delivery, revealed a mild degree of respiratory alkalosis with a compensatory metabolic acidosis. Umbilical cord blood gas results indicated the presence of significant fetal acidosis, both respiratory (mean pCO2 Uv 46,3 torr (SD 11,3) and Ua 54,3 torr (SD 12,0)), and metabolic (mean base excess Uv-9 mEq/l (SD 4,2) and Ua-11,8 mEq/l, (SD 5,0)) in origin. The average umbilical cord blood oxygen tensions were Uv 25,9 torr (SD 10), and Ua 15,4 torr (SD 8,5); mean maternal to fetal base-excess gradients were Ma-Uv 4,1 mEq/l (SD 2,8) and Ma-Ua 6,5 mEq/l (SD 3,5). Five mothers (10%) offered convincing evidence of factual recall during surgery, and 3 of these were aware of pain. Nausea and vomiting occurred in 5 patients and in 4 there were clinical signs of postoperative chest infection. The degree of fetal biochemical asphyxia, and the incidence of maternal awareness during surgery, were significantly greater than previously reported when thiopentone was used for the induction of anaesthesia for Caesarean section. The results obtained are discussed, and the conclusion is drawn that propanidid for anaesthesia appears to offer no advantage over thiopentone in obstetric practice.