The three principal methods of determining the electron temperature Te at ionospheric F-region heights are briefly reviewed and their relative merits discussed. Satellite observations suffer because of the difficulty in separately determining time and height dependencies, and rockets cannot be launched on a scale required to follow diurnal and seasonal changes. The radar backscatter technique has none of these disadvantages, but, in the case of the Millstone Hill radar, suffers from poor height resolution and difficulties arising in the interpretation of some of the observations. Results reported for the year 1963 from (a) rocket soundings from Wallops Island, Virginia, (b) the satellite Explorer 17, and (c) ionospheric backscatter observations from Millstone Hill, Westford, Massachusetts, are compared. These data are particularly well suited for intercomparison, since they all relate to roughly the same geographic location. There is only fair agreement between the height dependence of Te obtained by the backscatter and rocket techniques, but there is good agreement between the satellite and backscatter results for Te at a fixed altitude. It is believed that the limitations of the three techniques in part account for these differences, and hence the three methods complement one another well. All three techniques indicate that at midlatitudes the electron temperature exceeds the ion temperature both by day and by night. Differing estimates of the magnitude of the nighttime heat flux required to maintain this temperature difference make it difficult at present to establish the source. The backscatter results would, however, be consistent with the amount of heat expected to arise from the cooling of the protonosphere.