Thickness is an important parameter for the description of sea ice. Its spatial and temporal distribution modifies the interaction between ocean, ice and atmosphere. Knowledge of the ice thickness distribution is of particular importance for observations of climatic changes, validation of sea-ice and general circulation models as well as for human activities in the polar regions.So far, only few operational techniques for the determination of ice thickness distributions exist. Thus, there is still a great need for the development of accurate and simple, generally applicable methods.In this thesis, two geophysical methods, a seismic and an electromagnetic-inductive technique, are examined for their accuracy and general applicability. Both aspects are investigated by means of comparisons of drill-hole deter-mined with geophysi-cally derived thicknesses along extended profiles. Additionally, the thickness range, resolution and sensitivity against variable ice properties are examined by means of theoretical model calculations.The porosity of the ice and its electrical conductivity are derived by means of ice core analyses. These are the main variables influencing the propagation of elastic waves and the development of electromagnetic fields in the ice.With the seismic measurements ice thickness is determined from the dispersion of surface waves. On average, derived thicknesses underestimate the true thickness by about 20%. The lateral resolution is not better than 20 m. Due to its high porosity during summer the ice strength is very reduced. This considerably hampers the propagation of elastic waves such that ice thickness can hardly be determined during this season.Execution and analysis of the measurements are involved. Nevertheless, the seismic technique is the only one which yields information about bulk elastic parameters of larger ice areas. Here, such parameters are calculated from propagation velocities for summer and winter first- and multi-year ice.Electromagnetic induction measurements make use of the small electrical ice conductivity. Therefore, a primary electromagnetic field only induces eddy currents within the sea water below the ice. Thus, the thickness measurement is actually a determination of the distance between the sea water and the instrument. Over level ice with thicknesses up to 5 to 6 m, the deviations from drill-hole determined thickness at the same point are smaller than 10%. The lateral resolution is a few meters. In contrast, close to pressure ridges or over deformed ice the comparison with drill-hole measurements can show bigger deviations. Here, the electromagnetic (EM) measurements are difficult to interpret. This is a result of the large area in which eddy currents are generated below the instrument. Maximum thicknesses of pressure ridges are generally underestimated.The conductivity of Arctic ice is very low compared to that of sea water. It shows only small seasonal variation with little effect on the EM signal. This is also true for melt water accumulating on or within the ice during summer. The investigation therefore shows the general applicability of the EM technique in almost all Arctic regions and at any time.As is shown from measurements and theory, these results are only partially valid for measurements over Antarctic sea ice in summer.The simple principle and the great progress of EM measurements are utilised to perform continuous ice thickness measurements from an icebreaking ship. With these measurements different ice regimes can clearly be distinguished and mean ice thicknesses are derived. The additional use of a laser altimeter enables to simultaneously measure the ice-surface roughness with high resolution. The lateral resolution of the EM measurements is highly reduced due to a greater height of the instrument above the ice or water surface, a reduced depth penetration or sensitivity, respectively, and the long time constant of the instrument in use. Due to these limitations, the thickness of single pressure ridges cannot be determined with this instrument setup.Ice thickness distributions for the Bellingshausen- and Amundsen Seas as well as for the Laptev Sea are shown as examples for applications of the EM technique and the results are discussed.