Paleosols are formed when the topsoil gets buried by the lateral distribution of soil and can store large quantities of soil organic matter (SOM) that may persist over millennial timescales due to its detachment from the disturbances at the surface. We studied buried SOM dynamics in the Brady paleosol, a deep loess (aeolian) deposit in Nebraska, USA, where climate has historically driven varying rates of loess deposition during the late Pleistocene and Holocene, burying soils up to 50m below the surface. Soils were sampled along the depositional and erosional transects at depths from 0.2 to 5.5m to understand the variability in the physical and chemical composition of the soils. We used elemental and isotopic compositions of C, N, 13C, and 15N, along with radiocarbon, base cation concentrations, and Fourier Transformed Infrared Spectroscopy (FTIR) to determine the distribution, stabilization, and composition of SOM and organic carbon in the soil profiles. Our results show a general decreasing trend of 13C and 15N values with depth, suggesting root input to soil carbon pools and the presence of less decomposed SOM in the deeply buried soil layers. Radiocarbon analysis of bulk soil indicated a loss of ancient carbon and incorporation of new photosynthate carbon in the eroding transect. Our stabilization study indicated, modern and buried Brady soil both are flocculated in physical structure and Brady soils has more monovalent cations compared to the modern soils. To determine the vulnerability of the SOM to the addition of moisture, we added water to soil from the different transects and depths at 60% pore space capacity in two different experimental setups – repeated wet and dry cycles and continuously wet. We found that repeated wetting and drying led to higher CO2 efflux for buried Brady soils from erosional transects compared to the modern soils collected from the depositional transects. Our findings indicate that change in soil moisture status can play a critical role in destabilizing previously protected ancient carbon. Finally, our study highlights the need for furthering our understanding of how a predicted increase in precipitation quantity and intensity coupled with accelerated erosion can release large quantities of greenhouse gases by mineralization of previously protected old carbon stocks.