Most research on soil compaction and its influence on soil functions rely on the assumption of soil being a homogenous elastic material. Attenuation of stresses applied to an elastic material is easily described based on the loaded area and the stress distribution in that area. However, agricultural soils, and in particular regularly tilled topsoils, are very heterogenous. In the StressSoil project, we wanted to acknowledge these material properties and elucidate stress transmission, soil failure and the resulting impacts on soil functions without a priori assumptions. We hypothesized that mechanical strength of agricultural soils shows a scale-dependency related to soil structural units. This would mean that inter-aggregates bonds and aggregate strength are crucial in understanding the stress transmission patterns through agricultural soils as well as deformation. We formulated three hypotheses at the beginning of the StressSoil project: (i) inter-aggregate bonds are the relevant soil mechanical parameters for the definition of the soil failure criterion; (ii) inter-aggregate bonds follow a continuous probability distribution integrating their scale dependency; (iii) the parameters of the probability distribution describing mainly cohesion and friction will enhance the prediction of soil compaction.