Exercise-induced arterial hypoxaemia is defined as a reduction in the arterial O-2 pressure (PaO2) by more than 1 kPa and/or a haemoglobin O-2 saturation (SaO(2)) below 95%. With blood gas analyses ideally reported at the actual body temperature, desaturation is a consistent finding during maximal ergometer rowing. Arterial desaturation is most pronounced at the end of a maximal exercise bout, whereas the reduction in PaO2 is established from the onset of exercise. Exercise-induced arterial hypoxaemia is multifactorial. The ability to maintain a high alveolar O-2 pressure (PAO(2)) is critical for blood oxygenation and this appears to be difficult in large individuals. A large lung capacity and, in turn, diffusion capacity seem to protect PaO2. A widening of the PAO(2)-PaO2 difference does indicate that a diffusion limitation, a ventilation-perfusion mismatch and/or a shunt influence the transport of O-2 from alveoli to the pulmonary capillaries. An inspired O-2 fraction of 0.30 reduces the widened PAO(2)-PaO2 difference by 75% and prevents a reduction of PaO2 and SaO(2). With a marked increase in cardiac output, diffusion limitation combined with a fast transit time dominates the O-2 transport problem. Furthermore, a postexercise reduction in pulmonary diffusion capacity suggests that the alveolo-capillary membrane is affected. An antioxidant attenuates oxidative burst by neutrophilic granulocytes, but it does not affect PaO2, SaO(2) or O-2 uptake (VO2), and the ventilatory response to maximal exercise also remains the same. It is proposed, though, that increased concentration of certain cytokines correlates to exercise-induced hypoxaemia as cytokines stimulate mast cells and basophilic granulocytes to degranulate histamine. The basophil count increases during maximal rowing. Equally, histamine release is associated with hypoxaemia and when the release of histamine is prevented, the reduction in PaO2 is attenuated. During maximal exercise, an extreme lactate spill-over to blood allows pH decrease to below 7.1 and according to the O-2 dissociation curve this is critical for SaO(2). When infusion of sodium bicarbonate maintains a stable blood buffer capacity, acidosis is attenuated and SaO(2) increases from 89% to 95%. This enables exercise capacity to increase, an effect also seen when O-2 supplementation to inspired air restores arterial oxygenation. In that case, exercise capacity increases less than can be explained by VO2 and CaO2. Furthermore, the change in muscle oxygenation during maximal exercise is not affected when hyperoxia and sodium bicarbonate attenuate desaturation. It is proposed that other organs benefit from enhanced O-2 availability, and especially the brain appears to increase its oxygenation during maximal exercise with hyperoxia.