Soil water is an important limiting factor for restoring alpine meadows on the northern Tibetan Plateau. Field studies of soil‐water content (SWC), however, are rare due to the harsh environment, especially in a mesoscale alpine‐meadow ecosystem. The objective of this study was to assess the spatial variability of SWC and the temporal variation of the spatial variability in a typical alpine meadow using a geostatistical approach. SWC was measured using a neutron probe to a depth of 50 cm at 113 locations on 22 sampling occasions in a 33.5‐hm2 alpine meadow during the 2015 and 2016 growing seasons. Mean SWC in the study plot for the two growing seasons was 18.7, 14.0, 13.9, 14.3, and 14.8% for depths of 10, 20, 30, 40, and 50 cm, respectively, and SWC was significantly larger at 10 cm than at other depths. SWC was negatively correlated with its spatial variability, and the spatial variability was higher when SWC was lower. Thirty‐three sampling locations in this study plot met the requirement of accuracy of the central limit theorem. A Gaussian model was the best fit for SWC semivariance at depths of 10, 20, and 30 cm, and the spatial structural ratio was between 0.997 and 1, indicating a strong spatial dependence of SWC. The sill and range fluctuated temporally, and the nugget and spatial structural ratio did not generally vary with time. The sill was significantly positively correlated with SWC and was initially stable and then tend to increase with SWC. The nugget, range, and spatial structure ratio, however, were not correlated with SWC. These results contribute to our understanding of SWC spatial distribution and variation in alpine meadows and provide basic empirical SWC data for mesoscale model simulations, optimizing sampling strategies and managing meadows on the Tibetan Plateau.