In this study, we use combined geophysical, mineralogical, terrain, and hydropedological approaches from representative toposequences to assess the scaling behaviour of variability in two saline-sodic soil-landscapes in upland South Australia. The complex patterns of spatial and temporal soil variation observed in these landscapes are the cumulative result of several geomorphological, hydrological and geochemical processes acting together. The soil-landscape process at one scale has a feedback to action at another scale. Consequently, spatial and temporal variations occur over several levels of scale and resolution, principally: (i) molecular (nano-scale), (ii) profile (point scale), (iii) plot (100 m2), (iv) toposequence (50 to 20,000 m), (v) catchment (100 to 500 ha) and (vi) regional scales (100 to 500 km2). In this paper two study areas were selected in a high rainfall zone (680 mm) and a low rainfall zone (475 mm) to study regional digital predictions of land degradation in high value rain fed farming areas via “upscaling” approaches. This approach was used to: (i) refine existing large scale soil maps (e.g. ~ 1: 5,000 scale) and reveal in more detail the distribution of the complex soil-landscape patterns, and (ii) identify near-surface hydropedological patterns featuring preferential water flow pathways within the complex soil-landscape patterns.The multiscale soil investigation approaches applied included: (i) near-surface geophysical surveys of potassium concentration (K %) and thorium (Th ppm) from gamma radiometrics, ground apparent electrical conductivity (ECa), and volume magnetic susceptibility (ę), (ii) soil survey (e.g. structural morphology) and physicochemical data (e.g. mineralogy, chemistry), and (iii) terrain analysis. The newly acquired hydropedological understanding for each of the study area landscapes underpinned: (i) construction of graphic 3D whole-of-landscape conceptual toposequence models that both highlighted and communicated more widely hillslope processes (e.g. soil types, soil morphological and salt distributions), and (ii) the development of rules-based frameworks to digitally predict (upscale) the distribution of saline-sodic patterns in nearby, regionally repeating toposequences. The application of the newly acquired hydropedological knowledge for each study area landscape played a key role by providing the conceptual link between the various scales of soil investigation that were conducted, allowing, for example, the fine scale soil system patterns to be aggregated to support interpretation of soil processes identified in more coarsely scaled soil system patterns. The application of hydropedological principles were successful in this respect because water movement is the common thread that influences (and links) the majority of processes within the hillslope soil systems investigated.