The thermal effect of the Qinghai-Tibet Plateau (QTP) affects the regional and global weather and climate. Although some studies have evaluated the vertically integrated atmospheric heat sources (< Q1>) calculated from different reanalysis datasets, an evaluation of the vertical distribution of atmospheric heat sources (Q1) is lacking. In this study, data from 14 radiosonde (TK) sites in the eastern QTP were interpolated to 1o×1o grids using the objective analysis method. The summer Q1 calculated using gridded TK data based on the "inverse algorithm" was compared to those of ERA-Interim, ERA5, JRA55, and CRA40. The applicability of the four reanalysis datasets over QTP was discussed from the perspective of the vertical distribution of Q1. The results show that the vertical profiles of meteorological elements of the four reanalysis datasets correlate well with that of TK, with a correlation coefficient of 0.99. However, the vertical velocities (ω) of the four reanalysis datasets differ greatly, with ω_ERA5 correlating best with ω_TK and having the smallest root mean square error (RMSE). The meridional and zonal mean Q1 and its components (local temperature change, temperature advection, and vertical shear of potential temperature) calculated using ERA5 correlate best with those estimated using TK, and the RMSE is the smallest. The ω and Q1 calculated using high-resolution reanalysis data correlate better with those from TK and have a lower RMSE. Under the same spatial and temporal resolution, ω_ERA5 and Q1_ERA5 correlate best with those of TK, and the RMSEs are the smallest. Among the four reanalysis datasets, ERA5 is more suitable for Q1 analysis in summer on the Eastern QTP.