The electron temperature Te in the Earth's plasmasphere up to 10,000 km altitude, has been measured routinely by a thermal electron energy distribution instrument on board the EXOS D satellite. The measurements made during the years 1989–1995 have been analyzed to obtain altitude (1000–8000 km) profiles of Te for magnetic latitudes 0°–40°N at different times of the day. The profiles are compared with those computed by the Sheffield University plasmasphere‐ionosphere model, modified to include nonlocal heating due to trapped photoelectrons and an equatorial high‐altitude heat source. The photoelectrons are trapped at altitudes above 600 km, and the high‐altitude heat source is applied as energy input along the magnetic field lines with apex altitude greater than 600 km between ±10° magnetic latitude. The results for 8000 km altitude show that the modeled values of Te computed without photoelectron trapping and the high‐altitude heat source are much less than the mean measured values, 3700 K compared with 6500 K. Depending upon altitude and latitude, a photoelectron trapping of up to 100% is required to raise the modeled electron temperatures to the mean measured values. However, photoelectron trapping alone cannot account for the observed latitude variation of Te, which depends on altitude. Model calculations carried out with 50% photoelectron trapping and an altitude and local time‐dependent high‐altitude heat source reproduce the measurements, including the latitude variation; the heat source required for nighttime is about one fourth of that required for daytime. An equatorial high‐altitude heat source appears to be the only mechanism that can account for the measured values of Te; values in excess of 12,000 K have been measured. There are some quantitative differences between the measured and modeled temperatures at night in the lower plasmasphere (<2500 km), which could be caused, in whole or in part, by the inaccuracies of the nighttime measurements due to the low plasma densities.