Upscaling ground-based moisture observations to satellite footprint-scale estimates is an important problem in remote sensing soil-moisture product validation. The reliability of validation is sensitive to the quality of input observation data and the upscaling strategy. This letter proposes a model-based geostatistical approach to scale up soil moisture with observations of unequal precision. It incorporates unequal precision in the spatial covariance structure and uses Monte Carlo simulation in combination with a block kriging (BK) upscaling strategy. The approach is illustrated with a real-world application for upscaling soil moisture in the Heihe Watershed Allied Telemetry Experimental Research experiment. The results show that BK with unequal precision observations can consider both random ground-based measurement errors and upscaling model error to achieve more reliable estimates. We conclude that this approach is appropriate to quantify upscaling uncertainties and to investigate the error propagation process in soil-moisture upscaling.