To better understand how carbon and nitrogen mineralization are linked in soils, we conducted a long-term incubation experiment and compared carbon and nitrogen dynamics in the bulk soil and in soil fractions. Topsoil of a Rendzic Leptosol from a beech forest site near Tuttlingen, Germany, was separated into three particle size classes: sand (2000–20 μm), silt (20–2 μm), and clay (<2 μm). Bulk soil and particle size fractions were incubated in replicate, allowing periodic destructive sampling of triplicates at day 0, 14, 42, 84, 140, 210, and 280. We monitored CO2–C respiration, NH3–N emissions, nitrogen mineralization, pool sizes of total and salt extractable (0.01 M CaCl2) organic carbon and nitrogen, and microbial biomass carbon and nitrogen. The chemical composition of selected samples was further characterized by 13C-NMR spectroscopy. Fractionation did not influence carbon mineralization (∑ fractions ≈ bulk soil), which decreased in the order sand > clay > silt. The fractions respired between 10.4% (sand fraction), 8.8% (clay fraction) and 4.4% (silt fraction) of total soil organic carbon. However, nitrogen mineralization was affected by the fractionation procedure (∑ fractions < bulk soil) and followed the order clay > silt > sand. Fractionation increased the surface area and hence provided accessory mineral surfaces, which allowed new binding of especially nitrogen-rich compounds, in addition to ammonium fixation via cation exchange. As indicated by lower metabolic quotients, microbial carbon mineralization was more efficient in the bulk soil compared to the calculated sum of fractions. In the clay fraction, carbon mineralization rates, salt extractable organic carbon contents, and microbial biomass carbon and nitrogen contents declined strongly towards the end of the incubation. This indicates that in the clay fraction, organic carbon was not available for microbial degradation and that microorganisms were strongly carbon-limited causing a subsequent inhibition of nitrogen immobilization. Density fractionation revealed that organic matter in the sand fraction consisted mainly of particulate organic matter present as light material containing partly decomposed plant remnants. The organic matter in the clay fraction was mostly adsorbed on mineral surfaces. Organic matter in the sand and in the clay fraction was dominated by O/N-alkyl C indicating low recalcitrance, but the C/N ratio of organic matter narrowed with decreasing particle size. Our results suggest that carbon and nitrogen mineralization are decoupled in the mineral-associated fractions of the soil. The specific interactions of both carbon and nitrogen containing components with the mineral matrix strongly modulate the mineralization dynamics. Therefore, isolated considerations of C/N or alkyl C to O/N-alkyl C ratios of organic matter are insufficient as indicators for decomposition in plant residues. The combined consideration of C/N and alkyl C to O/N-alkyl C ratios provides a first impression about the degree of decomposition in plant residues. However, bioavailability in fractions where organic matter is mainly stabilized by spatial inaccessibility and by organo-mineral interactions cannot be explained by these ratios, but can be examined by an incubation approach.