New Findings: What is the central question of this study? Is Tibetan and Chinese highlanders' blood oxygen-binding affinity (P50) different from that of other populations (at altitude or sea level), and does Tibetan P50 relate to haemoglobin concentration and/or exercise capacity at altitude? What is the main finding and its importance? Tibetans and Chinese at 4200 m have slightly lower P50 than sea-level residents. During maximal exercise at 4200 m, reduced P50 does not enhance pulmonary gas exchange, impair systemic oxygen extraction or affect peak exercise capacity. Oxygen saturation measurements based upon forehead oximetry are sufficiently reduced during exercise at altitude (and accurate compared with those obtained from arterial blood by co-oximetry) to determine P50 reliably. High-altitude natives are challenged by hypoxia, and a potential compensatory mechanism could be reduced blood oxygen-binding affinity (P50), as seen in several high-altitude mammalian species. In 21 Qinghai Tibetan and nine Han Chinese men, all resident at 4200 m, standard P50 was calculated from measurements of arterial PO2 and forehead oximeter oxygen saturation, which was validated in a separate examination of 13 healthy subjects residing at sea level. In both Tibetans and Han Chinese, standard P50 was 24.5 ± 1.4 and 24.5 ± 2.0 mmHg, respectively, and was lower than in the sea-level subjects (26.2 ± 0.6 mmHg, P < 0.01). There was no relationship between P50 and haemoglobin concentration (the latter ranging from 15.2 to 22.9 g dl-1 in Tibetans). During peak exercise, P50 was not associated with alveolar-arterial PO2 difference or peak O2 uptake per kilogram. There appears to be no apparent benefit of a lower P50 in this adult high-altitude Tibetan population.