Preloading with fluid is recommended before regional block in labour. Low dose epidurals may produce less haemodynamic disturbance than traditional stronger solutions of local anaesthetics. Our aim was to compare the incidence of hypotension in normal labouring women who received a low dose epidural (0.1% bupivacaine 15 mL with fentanyl 2 microg microg.mL(-1)) with and without an i.v. crystalloid preload. Women with normal labours were randomised to the intervention group: no i.v. crystalloid preload (n = 85) and the control group: 7 mL mL.kg(-1) i.v. crystalloid solution before epidural injection (n = 83). Mean arterial pressure was recorded every 5 min for 30 min. There was no difference between the groups in mean decrease in mean arterial pressure and similar proportions of women showed falls in mean arterial pressure of 20% or greater (13% vs. 11%, risk ratio 1.2, 95% CI 0.54 to 2.8, P = 0.6). Blinded analysis by independent obstetricians revealed no differences in the fetal heart rate abnormalities (20% vs. 15%, risk ratio 1.3, 95% CI 0.67 to 2.7). A scientifically valid conclusion whether preloading is useful cannot be drawn from this study. This study suggests that about 350 participants in each group would be necessary to exclude a type 2 error in a future study.