Abstract: This study collected soil moisture (SM) observations from 100 in situ observations and compared them with widely used SM estimations, derived from the Essential Climate Variables, European Reanalysis (ERA)‐Interim, the Modern‐Era Retrospective analysis for Research and Applications, and Noah at spatial resolutions of 0.25° × 0.25°, 0.5° × 0.5°, and 1° × 1° during unfreezing and freezing periods. Results indicate that all estimations except ERA can depict the SM changes on the Himalayan‐Tibetan Plateau (HTP) well to some extent and Noah performs the best as a whole. In the Ngari region, however, SM estimations substantially overestimate SM observations. In the Naqu region, except ERA, SM estimations overestimate the SM amount during freezing periods but underestimate the SM amount during unfreezing periods. In Maqu, except ERA, SM estimations can characterize SM variations with substantial underestimation of SM during unfreezing periods, however, slight underestimation of SM during freezing periods. In space, the Modern‐Era Retrospective analysis for Research and Applications and Noah can evaluate the spatial distribution of SM better. The SM changes across most parts of the HTP are heavily influenced by precipitation changes. In addition, these changes in the western flank of the HTP are the combined result of snow and glacial melting and precipitation. Except in the western and northern regions, SM during unfreezing periods is usually higher than during freezing periods. Besides, evaluations can better describe SM changes at medium and large spatial scales than at small spatial scales. This study provides important basis for the selection of proper remote sensing and reanalysis SM data sets for the analysis of SM on the HTP.