The circulation of the brain has been the object of more speculation and study than any other part of the body. All present methods of estimating cerebral blood flow, however, leave much to be desired. Since, for many practical purposes we are not interested in the blood flow per se but in whether or not the oxygenation of the brain is adequate, we have attempted to approach the problem of quantitative brain oxygen tension directly. The prime consideration was "what" to measure that might justifiably be called "mean tissue oxygen tension." As pointed out by Bronk et al.1and Bloor et al.2the spatial resolution of the oxygen electrode is such that the absolute value obtained from the cortex varies greatly with its position relative to blood vessels. Other obvious problems arise in using the exposed cortex, such as pulsation, movement, variations in electrode pressure, etc. However,