We have used the technique of double-blind sequential allocation to quantify the minimum local analgesic concentration (MLAC) of extradural bupivacaine for women in early (median cervical dilatation 2 cm) and late (median cervical dilatation 5 cm) labour. The first bolus was 20 ml of the bupivacaine test solution. The concentration was determined by the response of the previous woman to a higher or lower concentration of bupivacaine according to up and down sequential allocation. Efficacy was assessed using a 100-mm visual analogue pain score (VAPS). The test solution had to achieve a VAPS of 10 mm or less to be judged effective. In early labour, the MLAC of bupivacaine was 0.048% w/v (95% confidence intervals (CI) 0.037-0.058% w/v), and 0.140% w/v (95% CI 0.132-0.150% w/v) in the late group. The MLAC of bupivacaine in late labour was greater by a factor of 2.9 (95% CI 2.7-3.2) compared with the MLAC in early labour (P < 0.0001, 95% CI difference 0.08-0.11). We conclude that advancing labour requires an increased concentration of extradural bupivacaine for pain relief.