Abstract Three independent temperature datasets have been analyzed for quantifying the influence of the 11-year solar cycle modulation of the UV radiation. The datasets used include: US rocketsondes, the OHP lidar, and the global temperature database made by the successive SSU on the NOAA satellites, adjusted and provided by the UK Meteorological Office. These measurements cover the upper stratosphere and the mesosphere, where the direct photochemical effect is expected. The improvement of the analysis compared to previous ones was possible because the overall quality and the continuity of many data series have been checked more carefully during the last decade in order to look for anthropogenic fingerprints and the one used here have been recognized as the best series according to their temporal continuity. The analysis of the different data set is based on the same regression linear model. The 11-year solar temperature response observed presents a variable behavior, depending on the location. However, an overall adequate agreement among the results has been obtained, and thus the global picture of the solar impact in the upper stratosphere and lower mesosphere has been obtained and is presented here. In the tropics, a 1–2 K positive response in the mid and upper stratophere has been found, in agreement with photochemical theory and previous analyses. On the opposite, at mid-latitudes, negative responses of several Kelvin have been observed, during winters, in the analyses of the datasets analyzed here. In the mesosphere, at sub-tropic and mid-latitude regions, we observe a positive response all the year round increasing by a factor of two during winter.