The food security–climate change nexus rapidly gains momentum. Soil degradation plays an important role in this context while dealing with, for example, the productive capacity of our soil resources or carbon sequestration for climate change mitigation. However, little has been done to assess the pristine soil conditions despite the fact that these provide the basis to put changes into context. Various methodologies have been developed to assess the global distribution of current soil conditions. We used the S-World methodology that was developed to generate global soil property maps for environmental modelling studies. Up till now, the S-World methodology assessed current soil conditions by disaggregating the Harmonized World Soil Database using detailed information on climate, topography, land cover, and land use. This study used the S-World methodology to derive global soil conditions under natural vegetation. A large number of natural areas around the globe were identified for which land cover, expressed by the Normalized Difference Vegetation Index, could be successfully correlated to environmental conditions such as temperature, rainfall, and topography. Using this relation in regression kriging, the vegetation index under natural conditions was derived for the entire globe. Subsequently, the S-World methodology was used to calculate the soil properties under natural land cover and absence of human land use. Soil property maps for natural and current conditions were compared and showed large local differences. The results indicate that there are major changes due to land cover and land use change and that these changes are concentrated on the globe. The results are the basis for future assessments on, for example, land degradation, food security, or the sustainable development goals.