For understanding and predicting rainfall–runoff processes in watersheds, soils and their hydraulic properties play a central role. Experimentalists observe and document hydric soil indicators in detail for more and more sites in various catchments. Modelers, on the other hand, try to break down natural process complexity into models that are based on simplified process descriptions. The challenge for both is to identify first-order controls of catchment hydrologic behavior, which helps to better understand the nonlinearity of natural systems. This chapter describes how both, experimentalists and modelers, can work together toward a better understanding and quantification of subsurface runoff processes. Specifically, this chapter addresses the following questions: (1) How are subsurface runoff processes represented in models of different complexity, ranging from simple conceptual ones to more complex physically based ones? (2) How can catchment-scale models be parametrized using point-scale measurements and existing model approaches originally developed for small scales (e.g. a soil column)? (3) Which information can be gained from soil surveying methods, including mapping approaches of hydric soil indicators? (4) Can decision schemes be useful to indicate dominant runoff processes in an objective way? Finally we describe the soft data concept as a possible way forward to enhance the dialog between experimentalists and modelers. Soft data refer to all kinds of qualitative or semi-quantitative information on pedologic and hydrologic processes and properties. These data can be made useful for modeling by applying fuzzy-logic-based functions to evaluate the degree of acceptance of model simulation outputs compared to experimentalists’ field experience.