Hypoxaemia may occur after hyperventilation with nitrous oxide during labour. The purpose of this study was to assess whether diffusion hypoxia is a contributory factor. Twenty-four parturients were randomly allocated to receive 50 or 70% nitrous oxide in oxygen. The median nitrous oxide inhalation time per contraction was 58 s and 33 s, respectively. The end-tidal carbon dioxide and the minute ventilation remained unchanged. The end-tidal oxygen concentration was lowest at 120s, reaching 15.4% in both groups. The oxygen saturation did not differ between the groups with a lowest median value of 96% before the start of nitrous oxide inhalation. Two parturients had episodes ofdesaturation. Both had low end-tidal oxygen concentrations in association with the desaturation but, as the end-tidal nitrous oxide concentrations were low, the desaturations could not be attributed to diffusion hypoxia.