Information regarding soils and their variability in different landscapes is increasingly sought after to improve decision-making regarding a wide range of global issues, such as agricultural production, climate change and the problems of environmental degradation. This information comes in the form of a group of soil properties, indicated by field observations or laboratory analyses (i.e. organic matter (OM), particle size, calcium carbonate, iron, pH, humidity, etc.), the list and the determination procedures were the subjects of standardization. This group of soil properties, also called “primary” soil properties, is what makes up current soil databases and the resulting soil maps. These primary soil properties are used as input data for pedotransfer functions to estimate properties called “functional” soil properties (i.e. available water content, structural soil stability, a pesticide’s adsorption coefficient, etc.). These functional properties are then used to help decision about soil management, and also as input parameters of crop models, carbon dynamics models, hydrological models and erosion models. The main factor limiting the use of these models is the lack or low density and accuracy of the determination of primary soil properties needed to organize them. In fact, the existing soil databases in the world are neither sufficiently comprehensive nor sufficiently precise to meet the demands of soil data, especially to organize agri-environmental models as mentioned above. Therefore, there is a major challenge to develop alternative methods for mapping soil properties over large areas, with high spatial resolution, while presenting acceptable cost implementation.