<p>Formation of mineral-associated organic matter (MAOM) is a decisive process in the stabilization of OM against rapid microbial decomposition and thus in the soils&#8217; role as global carbon (C) sink. Sorption experiments of dissolved OM (DOM) repeatedly showed that particularly mineral subsoils have a large sorption capacity to retain more C. However, there is also an increasing body of literature, revealing an increasing output of dissolved organic C (DOC) from soils. Here, we investigated into this paradox in forest soil under beech by a combination of a field labelling experiment with <sup>13</sup>C-enriched litter with a unique DO<sup>13</sup>C and <sup>13</sup>CO<sub>2</sub> monitoring, an in-situ C exchange experiment with <sup>13</sup>C-coated minerals, and batch sorption experiments.</p><p>Within two years of <sup>13</sup>C monitoring, only 0.5% of litter-derived DO<sup>13</sup>C entered the subsoil, where it was only short-term stabilized by formation of MAOM but prone to fast microbial mineralization. The <sup>13</sup>C monitoring, sorption/desorption experiments in the laboratory, and also the in-situ C exchange on buried soil minerals revealed that there is a frequent exchange of DOM with native OM and a preferential desorption of recently retained OM. Hence, there appeared to be a steady-state equilibrium between C input and output, facilitated by exchange and microbial mineralization of an adopted microbial community. The remobilized OM was also richer in less sorptive carbohydrates. Along with transport of most of DOM along preferential paths, this further increased the discrepancy between laboratory-measured sorption capacities of subsoil and the actual C loading of minerals. Finally, the <sup>13</sup>C labeling experiments revealed that input of fresh litter-derived OM into subsoil may even mobilize old-soil derived OM. Hence, in the field different biogeochemical constraints are acting that prevent that the laboratory-based C sink can be reached in the field.&#160; We conclude, that forest subsoils can hardly be considered as additional C sink, even at management options that increase DOC input to subsoil.</p>