During acclimatization to the hypoxia of altitude, the cerebral circulation is exposed to arterial hypoxia and hypocapnia, two stimuli with opposing influences on cerebral blood flow (CBF). In order to understand the resultant changes in CBF, this study examined the responses of CBF during a period of constant mild hypoxia both with and without concomitant regulation of arterial P(CO2). Nine subjects were each exposed to two protocols in a purpose-built chamber: (1) 48 h of isocapnic hypoxia (Protocol I), where end-tidal P(O2) (P(ET,O2)) was held at 60 Torr and end-tidal P(CO2) (P(ET,CO2)) at the subject's resting value prior to experimentation; and (2) 48 h of poikilocapnic hypoxia (Protocol P), where P(ET,O2) was held at 60 Torr and P(ET,CO2) was uncontrolled. Transcranial Doppler ultrasound was used to assess CBF. At 24 h intervals during and after the hypoxic exposure CBF was measured and the sensitivity of CBF to acute variations in P(O2) and P(CO2) was determined. During Protocol P, P(ET,CO2) decreased by 13% (P < 0.001) and CBF decreased by 6% (P < 0.05), whereas during Protocol I, P(ET,CO2) and CBF remained unchanged. The sensitivity of CBF to acute variations in P(O2) and P(CO2) increased by 103% (P < 0.001) and 28% (P < 0.01), respectively, over the 48 h period of hypoxia. These changes did not differ between protocols. In conclusion, CBF decreases during mild poikilocapnic hypoxia, indicating that there is a predominant effect on CBF of the associated arterial hypocapnia. This fall occurs despite increases in the sensitivity of CBF to acute variations in P(O2)/P(CO2) arising directly from the hypoxic exposure.