The paper by Dominelli et al. (2020) is a very interesting publication and we follow Jerome Dempsey's suggestion (Dempsey, 2020) to discuss it. The role of haemoglobin oxygen affinity for oxygen uptake during exercise in normoxia and hypoxia has been investigated and discussed for decades. The importance of the article is to show that a left shift of the oxygen dissociation curve (ODC) of high affinity haemoglobins (HAH) may be beneficial during exercise at moderate altitude (3000 m) not only in animals but also in humans, thus confirming the case report in siblings by Hebbel et al. (1978). Moreover, the surprisingly high arterial oxygen saturation in many Ethiopians (Beall et al. 2002) suggests that high affinity may also be caused by still-unknown effects, because a large number of modified Hb molecules are not known in this population (Cheong et al. 2017). These observations are in contrast to the right shift of the (in vitro!) standard ODC by upregulation of 2,3-biphosphoglycerate concentration usually observed in moderate hypoxia. However, at extreme altitude (above 5000 m) strong hyperventilation overruns this effect, thus hindering too large a decrease of arterial saturation (Samaja et al. 2003; West et al. 2013). A specially important point is not mentioned in the paper discussed here. Oxygen delivery to the tissues is influenced by not only the position of the ODC, measured as half-saturation pressure P50, but also its shape. According to Berglund (1972) and Boyer et al. (1972) Hill's n, i.e. the slope of the ODC in a logarithmic presentation, of the high affinity Hb Malmö is much smaller (1.50–1.58) than that of normal Hb (2.86). The stabilizing effect of the marked S-form of the normal dissociation curve on at low saturation is markedly reduced and therefore the high Hb concentration is especially necessary for a sufficiently high in the tissue capillaries. What about Hill's n in Dominelli‘s experiments? Was it normal? This essential aspect is not considered. Interestingly, n is increased in South American highlanders (Schmidt et al. 1990), pointing to an important role of this adaptation. Measurements of the ODC have been mostly performed in vitro; this is also the case in the paper discussed here. The conditions during measurement in the Hemox-Analyzer used by Dominelli et al. are, however, even more artificial: a small volume of red cells is dispersed in a high volume of buffer solution and CO2 is absent. In any case the in vivo curve is steeper because of the lower pH (Bohr effect) and higher temperature in the tissue vessels, especially during exercise. Additionally, the Bohr effect determined in vivo is bigger than in vitro (Braumann et al. 1982), which can be partly explained as follows. The Bohr effect induced by CO2 at low saturation (the condition in tissue capillaries) is larger than for lactic acid (Meier et al. 1974). Lactic acid, however, enters the erythrocytes only slowly. But it liberates CO2 from bicarbonate in the plasma. CO2 diffuses rapidly into the red cells, therefore the in vivo Bohr effect is mainly caused by CO2 (Böning et al. 1991, 2007). Under strenuous exercise conditions, CO2 therefore accumulates in the muscle capillaries and may exceed 80 mmHg (Stringer et al. 1994). Due to this quasi-closed system, the Bohr effect strongly modulates the in vivo ODC and must therefore be considered particularly under exercise conditions. Furthermore, various papers published since the 1970s (reviewed in Böning et al. 2014) suggest that unknown factors make the in vivo curve even steeper than expected. Possible causes are concentration changes of chloride, glutathione or glutamate in the erythrocytes. Different numbers of males and females in the HAH and normal subjects groups, respectively, increase scattering, because sex differences in P50, Bohr coefficient and temperature coefficient of the ODC have been observed (Böning et al. 1978). The large scattering seems also to be the cause of a lack of significant differences between groups in weight, body mass index and peak power. Measuring method: “After deoxygenation with nitrogen, the compressed air reoxygenation curve was measured at pH 7.6 and 37°C using a laboratory-developed protocol (Winslow et al. 1977)”. We emphasize again that pH 7.6 and lack of CO2 do not represent physiological conditions, especially not in the muscle capillaries. In our investigations we therefore modified the continuous Hemox-Analyzer method using a plasma-like buffer and adding CO2 to the equilibration gas (e.g. Schmidt et al. 1990). Furthermore, Winslow et al. did not use the Hemox-Analyzer in the cited paper; instead they used equilibration of undiluted blood samples in tonometers at various . A non-standardized pedalling frequency possibly explains differences in oxygen uptake between groups at low exercise intensity (Fig. 7C in Dominelli's paper). Gross efficiency markedly changes with pedal rate (Böning et al. 1984). During intense exercise in normoxia, the concentration of lactate rises as result of increased metabolism, not because of a lack of oxygen (e.g. Brooks, 2010). In acute hypoxia, however, there is an additional increase during submaximal exercise. The lactate paradox appears only after acclimatization (detailed description in West et al. 2013). The higher end-exercise lactate concentration in the HAH group is most easily explained as result of a higher relative intensity visible in the increased oxygen uptake per watt. Base excess is a measure of fixed acids like HCl or lactic acid in blood, not including the buffer Hb. Hb is only used as a measure of the buffer value of blood and is necessary for calculation of base excess. We hope that some of these questions may be answered in further discussions. Nevertheless, this paper provides new insights by using an innovative approach, which should be expanded to include more in vivo aspects. None declared. D.B. contributed to conception or design of the work; acquisition or analysis or interpretation of data for the work; drafting the work or revising it critically for important intellectual content; final approval of the version to be published; agreement to be accountable for all aspects of the work. W.S. contributed to conception or design of the work; acquisition or analysis or interpretation of data for the work; drafting the work or revising it critically for important intellectual content; final approval of the version to be published; agreement to be accountable for all aspects of the work. None.