Aims. We applied the quasi-thermal noise (QTN) method to Parker Solar Probe (PSP) observations to derive the total electron temperature (T_e). We combined a set of encounters to make up a 12-day period of observations around each perihelion from encounter one (E01) to ten (E10), with E08 not included. Here, the heliocentric distance varies from about 13 to 60 solar radii (R_⊙). Methods. The QTN technique is a reliable tool to yield accurate measurements of the electron parameters in the solar wind. We obtained T_e from the linear fit of the high-frequency part of the QTN spectra acquired by the RFS/FIELDS instrument. Then, we provided the mean radial electron temperature profile, and examined the electron temperature gradients for different solar wind populations (i.e. classified by the proton bulk speed, V_p, and the solar wind mass flux). Results. We find that the total electron temperature decreases with the distance as ∼R^−0.66, which is much slower than adiabatic. The extrapolated T_e based on PSP observations is consistent with the exospheric solar wind model prediction at ∼10 R_⊙, Helios observations at ∼0.3 AU, and Wind observations at 1 AU. Also, T_e, extrapolated back to 10 R_⊙, is almost the same as the Strahl electron temperature, T_s (measured by SPAN-E), which is considered to be closely related to or even almost equal to the coronal electron temperature. Furthermore, the radial T_e profiles in the slower solar wind (or flux tube with larger mass flux) are steeper than those in the faster solar wind (or flux tube with smaller mass flux). The more pronounced anticorrelation of V_p–T_e is observed when the solar wind is slower and located closer to the Sun.