Soil mixing by earthworms can have a large impact on the fate of nutrients and pollutants and on the soil's ability to sequester carbon. Nevertheless, methods to quantify earthworm ingestion and egestion under field conditions are largely lacking. Soils of the Fennoscandian tundra offer a special possibility for such quantifications, as these soils commonly lack burrowing macrofauna and exhibit a well-defined O horizon with low bulk density on top of a mineral soil with higher density. Since ingestion-egestion mixes the two soil layers, the temporal changes in the bulk density profile of such soils may be useful for estimating field ingestion rates. In this study, we applied a model for earthworm burrowing through soil ingestion to observed changes in soil densities occurring in a mesocosm experiment carried out in the arctic during four summers with intact soil. The earthworms present in the mesocosms were Aporrectodea trapezoides, Aporrectodea tuberculata, Aporrectodea rosea, Lumbricus rubellus and Lumbricus Terrestris (fourth season only). We show that changes in soil density profiles can indeed be used to infer earthworm ingestion rates that are realistic in comparison to literature values. Although uncertainties in parameter values were sometimes large, the results from this study suggest that soil turnover rates and endogeic earthworm soil ingestion rates in tundra heath and meadow soils may be as high as those reported for temperate conditions. Such large ingestion rates can explain observed large morphological changes in arctic soils where dispersing earthworms have resulted in complete inmixing of the organic layer into the mineral soil. Our approach is applicable to soil profiles with marked vertical differences in bulk density such as the soils of the Fennoscandian tundra where earthworms are currently dispersing into new areas and to layered repacked soil samples that are incubated in the field.