Methoxyflurane or trichloroethylene in concentrations of 0.2% or 0.1% used in anaesthesia for Caesarean section were compared. There were no differences in the acid-base state or the clinical condition of the infant at delivery, or in the incidence of maternal complications, which could be related to the choice of volatile agent. Either volatile agent in the 0.2% concentration was associated with an overall incidence of maternal awareness and unpleasant dreams of 2.5%, compared with 4.9% when the 0.1% concentration was used. The incidence of maternal nausea and vomiting was less in the 0.2% series. There was an apparent relationship between the induction-delivery interval and (a) the degree of neonatal acidosis and (b) the Apgar-minus-colour scores at one minute ((A-C)1). The correlation between those indices of increasing foetal depression and increasing length of the uterine incision-delivery interval was highly significant. There was no systematic relationship between the one-minute score for "colour" and the umbilical artery PO2.