The effect of adrenaline on the efficacy of extradural block and plasma bupivacaine concentrations was investigated in women undergoing elective (n = 40) and emergency (n = 40) Caesarean section. Patients were randomly allocated within these two groups to receive 0.5% bupivacaine 20 ml either plain or with adrenaline 1 in 200,000, as a single fractionated extradural injection. The elective plain group needed significantly more supplementary analgesia compared with the other three groups (P less than 0.05). In the elective group, plasma bupivacaine concentrations were significantly lower in the subgroup receiving extradural adrenaline than in the plain subgroup. This effect was not observed when comparing only those who received bupivacaine 100 mg. In the emergency group, there were no significant differences in plasma bupivacaine concentrations between the plain and adrenaline subgroups. Maximum plasma concentrations correlated significantly (P less than 0.0001) with dose of bupivacaine (mg kg-1). It is concluded that extradural adrenaline does not usefully reduce systemic absorption of 0.5% bupivacaine, but may improve its efficacy in extradural anaesthesia for elective Caesarean section.