Evaporites after burial by denser sediments tend to form pillows and diapirs (salt structures). The topographic and internal structural characteristics of those deformed evaporites are important contributions to understanding of how those bodies formed by deformation and are the focus of this thesis. Models of salt deformation differ in their approach towards explaining how salt structures form. In some models, salt structures develop with a characteristic wavelength related to the pre-deformation salt thickness. Other models predict that the wavelength is not uniform but varies because of plastic behaviour of the burden, which may inhibit diapirism during its early stages. These models were constructed assuming that salt displacement results purely from instability of a less dense salt layer overlain by denser overburden, but salt deformation is also affected by extension and compression, and by internal properties of the salt layer (composition, trace water, grain size and temperature). Using 3D seismic and well data from 6 quadrants of the Dutch sector of the Southern North Sea, we have developed the following three analyses of salt structure topography and internal structure that can be used to help assess the suitability of the different models. First, an attempt was made simply to establish relationships between original salt thickness and wavelength of salt structures. Original salt thicknesses were estimated by spatially filtering the present thicknesses, values that are minima as loss by erosion and dissolution cannot be ruled out. Salt thickens towards the Central Graben in the north and thins over the Texel-Ijsselmeer High. After estimating spacing between salt structures in two dimensions, ratios of 12-20 between wavelength and original thickness were derived. These values exceed previous values reported for a UK North Sea pillow province, as expected from how pillows evolve into diapirs with progressive halokinetic deformation. The findings favour models in which overburden thickness and mechanic behaviour affect diapirism. Second, spectral analysis of salt topography derived from various published seismic data of pillows revealed fractal-like geometries. Fractal surfaces have no single dominant wavelength, rather a range of wavelengths is present. Seismic and well data in the Cleaverbank Platform allowed us to explore explanations for the fractal geometry, including varied loading by overburden, viscosity and density of overburden, and diapirism over basement faults and dykes. The fractal-like topography appears to have arisen from some of these factors combined. Third, the internal composition of evaporites was characterised from lithological data from wells located in different parts of pillow structures. Transition probability tables representing changes in lithology up well sequences were constructed. They reveal the original depositional sequence better preserved within areas between pillow crests and rim synclines. Layers of less mobile anhydrite and carbonates occur more commonly above the more mobile halite in rim syncline than in crest wells. These studies illustrate how numerical analysis of salt topographic and categorical data may support interpretation.