Abstract. Soil moisture and soil temperature (SMST) are important state variables for quantifying the exchange of heat and water between land and atmosphere. Yet, long-term, regional-scale in situ SMST measurements are scarce on the Tibetan Plateau (TP), with even fewer are available for multiple soil depths. Tibet-Obs is such a long-term, regional-scale SMST observatory in the TP that has been established 10 years ago and includes three SMST monitoring networks, i.e. Maqu, Naqu, and Ngari (including Ali and Shiquanhe), located in the cold humid area covered by grassland, the cold semiarid area dominated by tundra, and the cold arid area dominated by desert, respectively. This paper presents a long-term (∼ 10 years) SMST profile dataset collected from the Tibet-Obs, which includes the original in situ measurements at a 15 min interval collected between 2008 and 2019 from all the three networks and the spatially upscaled data (SMups and STups) for the Maqu and Shiquanhe networks. The quality of the upscaled data is proved to be good, with errors that are generally better than the measured accuracy of adopted SMST sensors. Long-term analysis of the upscaled SMST profile data shows that the amplitudes of SMST variations decrease with increasing soil depth, and the deeper soil layers present a later onset of freezing and an earlier start of thawing and, thus, a shorter freeze–thaw duration in both the Maqu and Shiquanhe networks. In addition, there are notable differences between the relationships of SMups and STups under freezing conditions for the Maqu and Shiquanhe networks. No significant trend can be found for the SMups profile in the warm season (from May to October) for both networks that is consistent with the tendency of precipitation. A similar finding is also found for the STups profile and air temperature in the Shiquanhe network during the warm season. For the cold season (from November to April), a drying trend is noted for the SMups above 20 cm in the Maqu network, while no significant trend is found for those in the Shiquanhe network. Comparisons between the long-term upscaled data and five reanalysis datasets, including the ECMWF reanalysis v5 (ERA5), Modern-Era Retrospective analysis for Research and Applications, version 2 (MERRA-2), Global Land Data Assimilation System version 2 Catchment Land Surface Model (GLDAS-2.1 CLSM), GLDAS-2.1 Noah, and GLDAS-2.1 variable infiltration capacity (GLDAS-2.1 VIC), indicate that none of the current model-based products can reproduce the seasonal variations and interannual trend changes in the measured SMST profile dynamics in both networks. All the products underestimate the STups at every depth, leading to an earlier onset of freezing and a later onset of thawing, which essentially demonstrates that the current models are not able to adequately simulate winter conditions on the TP. In short, the presented dataset would be valuable for evaluation and improvement in long-term satellite- and model-based SMST products on the TP, enhancing the understanding of TP hydrometeorological processes and their response to climate change. The dataset is available in the 4TU.ResearchData repository at https://doi.org/10.4121/20141567.v1 (Zhang et al., 2022).