The difference between some of the REG observations presented regarding CO2 inhalation and hyperventilation and others previously reported is probably due to the use of 28 CM2 area monopolar electrodes. They considerably reduce extracranial impedance effects as compared with the commonly used 4-CM2 electrodes. Therefore, it is not appropriate to compare the data presented with the data obtained using smaller electrodes. In addition, use of the monopolar configuration allows better localization of cerebral impedance changes than the bipolar configuration, though a bipolar configuration could also have been used for the present study on the systemic effect of various gas tensions. The animal experiments prove that the pulsatile change in tissue blood volume is the source of the impedance change. This was seen in the perfused head where the REG followed stroke volume, and not pressure or flow velocity, while continuous flow abolished the REG. It was shown that in the perfused hind leg the impedance of the perfusate influenced both total muscle impedance and impedance pulse amplitudes, while pulse pressure and flow velocity were kept constant. In human beings, mean REG changes correlated with cerebral blood flow changes during inhalation of 5 percent CO2 and 95 percent O2 and during hyperventilation. There was a similar correlation between mean REG and arterial pCO2.