Surface soil temperature is an important input parameter to a variety of environment models, such as global circulation models, radiative transfer, and land surface process models. Soil temperature is especially important for normalizing microwave radiobrightness temperatures in inverse radiative transfer modelling for soil moisture and vegetation optical depth retrieval. To ensure maximum accuracy of soil moisture retrieval models on a regional or global scale, spatially averaged temperature data are necessary. Since the variability of surface temperature in time and space is extremely high due to incoming solar radiation, air temperature, vegetation, soil physical properties, and topography, an aggregation of a few point measurements rarely provides a good spatial average. Remote sensing methods typically provide spatially averaged values needed. Thermal infrared sensors (TIR) measure the skin temperature, but usually require some atmospheric correction, and during periods of cloud cover they become unusable. Microwave sensors also have the potential for providing reliable estimates of spatially averaged soil temperature. Microwave instruments are also much less affected by atmospheric conditions and thus require little or no correction. A technique to estimate the effective temperature with vertical polarized high-frequency microwave brightness temperatures is presented. Calibration procedures with field observations are discussed, and a technique to estimate the soil temperature at the soil moisture sampling depth for 6.6 GHz is shown.