The motorization of ground penetrating radar systems based on grid-free positioning systems using Realtime Kinematic GPS enables the efficient large-scale application of multi-antenna arrays, covering up to five hectares per day while simultaneously reaching a sampling resolution of less than 10 cm cross-line spacing. The resulting novel three-dimensional high-resolution data sets provide new ways for the non-invasive investigation of archaeological landscapes. However, a considerable part of the geophysical data sets can be devoid of archaeological structures. Yet, these ``empty’’ areas can carry significant information about the palaeo-environmental dynamics of the study area. Palaeo-environmental features should be considered of equal importance for the comprehensive understanding of archaeological landscapes. Currently, most geophysical prospection surveys disregard this potential and often only focus entirely on the apparent archaeological features and information. One of the reasons for this deficit is based on the complexity of palaeo-environmental information and the specific requirements for its extraction, visualization and implementation into the interpretation process. Many palaeo-environmental features display a continuous, four-dimensional character, such as finely laminated beach deposits accumulated over time or laterally migrating palaeo-channel systems. Specifically these properties demand a three-dimensional visualization as fundamental basis for any further analysis and subsequent interpretation. The study presented is investigating the specific requirements for extraction and analysis of palaeo-environmental information contained in large-scale, high-resolution GPR data sets with particular regard to technical aspects of three-dimensional data visualization. For this purpose, two large-scale, high-resolution GPR data sets have been chosen from Viking Age sites in Vestfold, Norway, displaying different types of palaeo-environmental features such as beach deposits, coastal ridges, bedrock outcrops and extensive palaeo-channel systems. Data sets were acquired using a 16 channel 400 MHz MALÅ Imaging Radar Array system with 8 cm cross-line spacing, as well as a six channel 500 MHz Sensors & Software PulseEkko Pro SPIDAR system with 25 cm cross-line spacing resulting in very densely sampled data and high imaging resolution. Several 3D visualization and GIS software packages were tested with particular attention paid to their performance in handling large data sets, visualization modes and possibilities, data interoperability and eventually data integration with traditional archaeological interpretation results. Test results were incorporated into a work-flow for extraction, visualization and analysis of high-resolution GPR data sets for palaeo-environmental information. The experience gained with the newly developed workflow and archaeological prospection approach is presented.