Although sea ice thickness is not directly measured by means of radar sensors, its magnitude and variations are to a certain degree reflected by the sea ice surface characteristics. Establishing simple but robust empirical relationships between sea ice surface properties and SAR signatures is an alternative to a much more complex theoretical microwave modelling approach for sea ice mapping. In this study correlations between helicopter-borne sea ice thickness and roughness measurements and SAR signatures are investigated. Preliminary results from a comparison of a Radarsat-1 scene and sea ice thickness data are presented. The profiles and histograms show a good agreement between SAR backscatter and sea ice thickness, while the correlation coefficient of R 2 = 0.2 indicates a very poor relationship. One source of error considered as the reason for the poor correlation is a remaining co-registration error due to insufficient ice drift data and the immanent sensor accuracies. But as the main source of ambiguity the variety of different scattering mechanisms typical for multiyear ice is discussed. Further analysis of the thickness and the laser roughness data in conjunction with the SAR profiles is planned to identify roughness parameters that are connected to the backscattering behaviour of multiyear sea ice surfaces and indirectly to its thickness.