Subsoil contains about half of the soil organic carbon (C) at the global scale. However, C and nitrogen (N) in the subsoil are assumed to be relatively stable, and their response to agricultural management practices is uncertain. This study compared the effect of 20 different management systems, combining two tillage regimes (inversion tillage vs minimum tillage), two crop residue managements (retained vs harvested) and five different fertilizer sources (liquid swine manure, liquid dairy manure, solid poultry manure, complete mineral fertilizer (NPK) and a no-N control (PK)) on C and N stocks of two Brunisols with contrasting textures (sandy loam and silty clay). We compared C and N stocks of multiple soil layers (0–15, 15–30, 30–45 and 45–60 cm depths) on the 8th year of this field trial. Overall, stock differences among management systems were relatively small in the top 15 cm, varying from 0.36 to 0.76 kg C m−2 and 0.01–0.04 kg N m−2. In subsurface soil layers (15–30, 30–45 and 45–60 cm), however, management practices considerably influenced C and N stocks (differences of up to 3.9 kg C m−2 and 0.3 kg N m−2 between treatments) in ways different than those observed in the topsoil. In the subsurface (> 15 cm) layers of the sandy loam soil, systems combining inversion tillage, residue retention, and liquid animal manure (especially swine manure) resulted in the greatest C and N stocks. Conversely, in systems combining inversion tillage and mineral fertilizer, C and N stocks in the subsurface layers were higher when residues were harvested than when they were not. Subsurface C and N stocks therefore appeared to be particularly responsive to different combinations of fertilizer sources and residue management practices under inversion tillage in the coarse-textured sandy loam soil. In the silty clay, irrespective of tillage treatments, N stocks were the highest in the 30–45 cm layer when residues were harvested, and liquid swine manure was used. Ultimately, management-differences in subsurface C and N stocks were so large compared to those in the topsoil, that they dominated the response of the whole soil profile to management practices in both soil types. These results demonstrate that subsurface C and N dynamics must be considered and that recommendation for C stock management in agricultural soils should include deep soil layers, at least under certain pedoclimatic conditions.