Objective: Potential effects of extradural analgesia on the progress of labour and obstetric outcome are still a matter of concern and the focus of ongoing debates. Despite this, little attention is paid to the initiation of extradural labour analgesia. The objective of this prospective, randomized, double-blind trial was to identify the optimal of three concentrations of bupivacaine used as a loading bolus for initiating extradural analgesia during labour. Methods: Sixty-seven full-term parturients requesting extradural analgesia received either bupivacaine 0.25% (group A), 0.125% (group B) or 0.0625% (group C). Bupivacaine administration was titrated to achieve a pain score ≤1.5 on a 10-cm visual analogue scale (VAS) and was limited to a maximum of 20 ml of the study drug. We recorded the time to achieve a VAS ≤1.5, the total amount of local anaesthetic used, sensory level, maternal and foetal as well as neonatal side effects. Results: The probability of obtaining a VAS pain score ≤1.5 was significantly lower in group C. Although mean times to reach a VAS pain score ≤1.5 were not statistically different, survival curves show a significantly lower probability for reaching this score. Significantly more parturients in group A developed a motor block ≥1 based on using a modified Bromage scoring scale. There were no differences in incidence of maternal hypotension, ephedrine requirements, foetal heart rate abnormalities, mode of delivery and neonatal outcome. Conclusions: Of the three concentrations used in this clinical setting, 0.125% was the most suitable concentration of plain bupivacaine to initiate extradural analgesia in labour. Using 0.25% bupivacaine increased the incidence of motor block, whereas for 0.0625% plain bupivacaine the probability to achieve adequate analgesia was unacceptably low.