The aim of this study was to determine whether permitting women in labour to eat a light diet would: (i) alter their metabolic profile, (ii) influence the outcome of labour, and (iii) increase residual gastric volume and consequent risk of pulmonary aspiration. Women were randomised to receive either a light diet (eating group, n = 48) or water only (starved group, n = 46) during labour. The light diet prevented the rise in plasma beta-hydroxybutyrate (p = 2.3 x 10(-5)) and nonesterified fatty acids (p = 9.3 x 10(-7)) seen in the starved group. Plasma glucose (p = 0.003) and insulin (p = 0.017) rose in the eating group but there was no difference in plasma lactate (p = 0.167) between the groups. There were no differences between the groups with respect to duration of first or second stage of labour, oxytocin requirements, mode of delivery, Apgar scores or umbilical artery and venous blood samples. Relative gastric volumes estimated by ultrasound measurement of gastric antral cross-sectional area were larger (p = 0.001) in the eating group. This was supported by the observation that those from this group who vomited, vomited significantly larger volumes than those in the starved group (p = 0.001). We conclude that eating in labour prevents the development of ketosis but significantly increases residual gastric volume.