This paper describes the CryoSat satellite mission, due for launch in 2005, whose aim is to accurately determine the trends in Earth’s continental and marine ice fields. The paper’s purpose is to provide scientific users of the CryoSat data with a description of the design and operation of the SIRAL radar and the CryoSat platform, the data products, and the expected error budget. The ‘low-resolution mode’ (LRM), ‘synthetic aperture mode’ (SARM) and “synthetic aperture interferometric mode’ (SARInM) of the SIRAL radar are described, together with its system parameters, its antenna gain pattern and interferometer phase difference pattern, and its calibration modes. The orbit is described, together with the platform attitude and altitude control law and control systems, and the expected pointing and altitude knowledge. The geographical masks that are used to determine acquisitions in the three SIRAL modes are described. The SIRAL data products, and the processing applied to produce them, are described. Level 1b, level 2 and higher-level products are described in turn, with a particular emphasis on the new procedures applied to the SARInM and SARM processing over ice surfaces. The beam forming and multi-looking is summarised, and a description is given of the behaviour of the SARM and SARInM echoes over idealised surfaces. These inform descriptions of the elevation retrievals of the level 2 processing, including the SARInM retrieval of interferometric phase. The combination of these data, through cross-over analysis over continental ice sheets, and through averaging over sea-ice, to determine areal averages of ice sheet elevation change or sea-ice thickness, is described. The error budget in these higher-level products is described, together with its breakdown into errors arising from the instrument and errors arising from the retrievals. The importance of the co-variance of these errors in determining the final error is stressed. The description of the errors also includes a summary of the experiments required following the launch to validate the CryoSat mission data. An estimate of the mission performance over ice surfaces is made at various spatial scales, and it is concluded that even the relatively short, three-year duration of the CryoSat mission will allow it to make an important scientific contribution, particularly when combined with results from earlier satellite missions.