An analysis and evaluation are presented of a technique for extracting infromation on neutral thermospheric temperature form satellite measurements of the relative abundances of H+ and O+ ions in the F region. Charge-exchange equilibrium in the neighborhood of 320 km relates the observed altitude distribution of this ion concentration ratio directly to that of the associated neutral species and thus to the neutral temperature. The method has been applied to an extensive set of ion concentration data acquired with the Atmosphere Explorer C and D satellites, and evaluated through direct comparison of these 'ion-ratio' temperatures with other concurrent in situ measurements. A comprehensive orbit-by-orbit comparison is presented with both ion temperature data and with calculations from the MSIS model. Select comparisons are also available with direct temperature measurements from the Nate instrument, and with temperatures derived from height distributions of observed neutral particle densities. Statistical analyses emphasize random and systematic differences between these several alternative types of temperature data. Sources and estimates of possible uncertainties unique with the ion-ratio method are also discussed. Excellent agreement has been achieved between the ion-ratio and the ion temperatures from the AE-D satellite; the analysis is, however, unable to account for a small systematic difference between corresponding measurements from AE-C. A random uncertainty of ~5% must also be associated with neutral temperatures derived from the ion-ratio method. Otherwise, a general consistency is demonstrated between the various forms of temperature data. All classes of measurements indicate that nighttime thermospheric temperatures are lower than predicted by the MSIS model.