The origins and development of early medieval towns in the Low Countries have been the object of long-standing discussion. This project aims to bring new information to this debate by examining a source of data that has been overlooked for a long time. So-called "dark earths" are often found in central places of which the origins are under debate. They are homogeneous, humus-rich units, often encountered between Roman strata and those of medieval towns. Since they are very difficult to study, they have often been discarded without examination or merely sieved, losing all contextual and stratigraphical relationships. However, these units represent the actual archaeological relics of the activities that can be related to the formation of early medieval centres. New, scientific techniques such as micromorphology (the analysis of undisturbed soil samples under a microscope) allow us to study these units without losing contextual information. In this project the dark earths from three Belgian sites (Antwerp, Tongeren and Lier) will be studied and compared in an international context in order to understand the human and natural processes that impacted the origins of (early) medieval towns, and how they evolved out of early central places, trade centres or Roman antecedents. By looking at these soils from a microachaeological perspective, new data are generated that enable us to interpret how and why medieval towns were formed in their social and spatial setting.