A retrieval algorithm for soil moisture within the uppermost metre of soil is presented. As calibration data, longtime soil moisture measurements from the former Soviet Union are used. The retrieval works in two steps. First, the distribution of longtime mean soil moisture is derived by using precipitation, soil texture, vegetation density and terrain slope. In a second step, the temporal variability at each location is deduced by using microwave radiation measurements available from satellite together with precipitation and air temperature data. This soil moisture algorithm is applied in Northern and Central Europe to validate a climate simulation from the regional model REMO.