Estimation of surface soil moisture is one of the major potential applications of radar remote sensing. The European Remote Sensing Satellites (ERS1 and 2) are equipped with a synthetic aperture radar working at C-band (5 Ghz) using a rather low incidence angle (23°). For this fequency and angle, the effect of soil roughness and vegetation attenuation are not negligible. It is shown that, for wheat canopy, it is possible to apply an empirical relation for correcting for the effect of vegetation. The proposed algorithm is derived from a data set acquired over several years using an airborne radar. It uses a simple cloud model to describe the vegetation attenuation. This algorithm does not need very precise information on vegetation density and yields a final precision for the moisture content on the order of 0.05 cm3/cm3.