High altitude (HA) dwellers have an exceptionally high tolerance to orthostatic stress, and this may partly be related to their high packed cell and blood volumes. However, it is not known whether their orthostatic tolerance would be changed after relief of the altitude-related hypoxia. Furthermore, orthostatic tolerance is known also to be influenced by the efficiency of the control of peripheral vascular resistance and by the effectiveness of cerebral autoregulation and these have not been reported in HA dwellers. In this study we examined plasma volume, orthostatic tolerance and peripheral vascular and cerebrovascular responses to orthostatic stress in HA dwellers, including some with chronic mountain sickness (CMS) in whom packed cell and blood volumes are particularly large. Eleven HA control subjects and 11 CMS patients underwent orthostatic stress testing, comprising head-up tilting with lower body suction, at their resident altitude (4338 m) and at sea level. Blood pressure (Portapres), heart rate (ECG), brachial and middle cerebral artery blood velocities (Doppler) were recorded during the test. Plasma volumes were found to be similar in both groups and at both locations. Packed cell and blood volumes were higher in CMS patients than controls. All subjects had very good orthostatic tolerances at both locations, compared to previously published data in lowland dwellers. In CMS patients responses of forearm vascular resistance to the orthostatic stress, at sea level, were smaller than controls (P < 0.05). Cerebral blood velocity was less in CMS than in controls (P < 0.01) and, at sea level, it decreased more than the controls in response to head-up tilting (P < 0.02). Cerebral autoregulation, assessed from the relationship between cerebral pressure and velocity, was also impaired in CMS patients compared to HA controls, when examined at sea level (P < 0.02). These results have shown that the good orthostatic tolerance seen in high altitude dwellers at altitude is also seen at sea level. There was no difference in orthostatic tolerance between CMS patients, with their exceptionally large blood volumes, and the HA controls. This may be because peripheral vascular and cerebrovascular responses (at least at sea level) are impaired in the CMS patients relative to HA controls. Thus, the advantage of the large blood volume may be offset by the smaller vascular responses.