Recent evidence suggests that cytochrome oxidase is partially reduced under resting conditions in the brain. Previous data, recorded over the past 30 years from intact brain using optical methods in the visible wavelength range, are consistent with this observation. These older data, while not conclusive in themselves, support the overall conclusions. The historical perspective on the experiments and controversies illustrates a number of useful principles. The first is that new methods tend to produce new observations that may be difficult to reproduce due to the uniqueness of the instrumentation. The second is that any new and different observations cannot be assimilated without an acceptable theoretical framework and, without assimilation can have little impact. Finally, the mechanisms which might explain why cytochrome oxidase may be more reduced than previously thought are still not fully developed and, therefore, the physiological significance of such reduction is not known.