Neutral atomic oxygen (AO)—the dominant atmospheric species at typical low Earth orbit altitudes—is responsible for the erosion, or other degradation, of many satellite materials. Therefore, AO has become an important consideration for spacecraft designers and manufacturers. The study of AO is also of interest to atmospheric physicists because it is involved in many of the chemical reactions occurring naturally in the mesosphere and lower thermosphere. Both these groups rely on atmospheric models for computer-based simulation and prediction of atomic oxygen concentrations. Such models require, or are enhanced by, empirical input data—that is, actual measurements of AO number densities. A review is presented of the different measurement techniques that, to date, have been used on satellites and sounding rockets to perform AO studies. Rather than reporting results from every sensor application, this article takes a more general view of the experimental methods, using example devices to highlight their advantages and disadvantages. New or promising equipment, or techniques that could be exploited for performing such measurements, are also described. We attempt some semiquantitative comparison of the techniques, although the most appropriate experimental method for any given flight opportunity depends heavily on the mission conditions and science goals. Our emphasis is on missions where the available mass and power are limited. In these situations the most suitable established device is probably that of the thin film actinometer. If more risk can be assumed then a more promising, but as yet unqualified, method is that of the fiber-optic reflectance sensor. However, since both these devices are nonreusable, it is shown that semiconducting sensors may be better for long duration, mass- and power-limited applications.