Satellite altimetric measurements by the 2010 launched ESA satellite CryoSat-2 are capable of obtaining the thickness distribution of marine ice fields. With its on-board Ku- band radar altimeter we retrieve the sea-ice freeboard, the height of the ice surface above the actual sea level, and finally the thickness by assuming hydrostatic equilibrium. In this thesis I estimate CryoSat-2 Arctic sea-ice freeboard and thickness and their corresponding uncertainties. In particular I focus on the impact of different retracking algorithms that are used to retrieve CryoSat-2 range estimates. In recent studies, snow is considered as transparent for Ku-band frequencies, although a possible bias is acknowledged. Therefore, another goal of this work is to investigate the impact of snow coverage since it may affect the backscatter of the radar signal due to its physical properties. Likewise we use validation measurements from airborne laser altimetry, ice-mass balance buoys and in-situ drilling to evaluate CryoSat-2 sea-ice retrievals on both hemispheres. The findings show that random uncertainties are dominated by speckle noise and the density of leads along the flight track of the satellite. On the other hand, systematic uncertainties result from the choice of the retracking algorithm and in particular the echo power threshold which is used to determine the main scattering horizon. This bias is accompanied by systematic uncertainties due to scattering within the snow layer in conjunction with surface roughness effects. Moreover, we find observational evidence that CryoSat-2 freeboard increase correlates with snow accumulation events over Arctic multiyear ice, regardless of the used retracking algorithm. Such biases may lead to overestimated sea-ice thickness, as observed in autumn 2013 north of Canada. However, comparisons with airborne laser altimetry data, in-situ drilling measurements and other remote sensing sensors show that the freeboard and thickness retrieval represent the geographical distribution of sea-ice types. In the future, a data fusion with thickness retrievals from other past, present and future satellite sensors has the capability to reduce the uncertainty level and enables the prediction of long-term trends in sea-ice volume.