Abstract. In order to improve our knowledge of the emitted signal of forests at L-band (1.4 GHz) we focused this study on permittivity measurements of heterogenic natural media such as soil or litter consisting of plant debris and organic matter. This study was done in the context of the upcoming SMOS (Soil Moisture and Ocean Salinity) satellite mission that will attempt to map surface soil moisture from L-band (1.4 GHz) passive microwave measurements. In the field of passive microwaves, very little information exists about the behaviour of the L-band signal of forests especially when litter is included in the soil-vegetation system. To date very few analyses have investigated the dielectric behaviour of the litter layer and its influence on the microwave emission of forests is generally neglected. This paper describes a technique used to measure the permittivity of soil and litter with a rectangular waveguide taking into account the heterogeneity of the media. The study was based on soil and litter samples collected in the coniferous forest of Les Landes, near Bordeaux, France. Particular attention was given to the impact of measurement errors on the calculation of permittivity. The measurements of the soil and litter dielectric constants were then used in a numerical model to simulate the microwave emissivity of the soil-litter system. As an example, it is shown that error in measurements lead to variations in the emissivity calculations of up to 0.025 at maximum (for an emissivity of 0.73, SM = 30%), which equals a total error of 3.36% of the calculated soil-litter emissivity.It was found that variations in the litter dielectric constant and layer thickness have a strong influence on the calculation of forest ground emissivity. As an example a 3 cm litter thickness can lead to a change of ∼0.15 between the soil and the soil-litter system emissivity (which equals an increase of about 20.4% in terms of surface emissivity).