AbstractIn soil organic carbon (SOC) survey, the role of sampling approach is crucial and should not be underestimated. In this sense, the appropriateness of soil sampling by pedogenic horizons (PGH) and fixed depth layers (FDL) in forest is still a field of debate. The present work aimed to: (a) study the spatial variability of SOC concentrations and C stocks through PGH and FDL samplings in uneven‐aged sweet chestnut, European beech and Norway spruce forests of the Apennine chain (Italy); (b) clarify the major advantages and drawbacks of sampling by PGH and FDL. In a representative area (18 × 18 m) of each forest, a soil profile was investigated and eight additional sampling points were selected. The sampling was performed both by PGH and FDL (0–15 and 15–30 cm). For each forest, no SOC content and C stock differences in 0–30 cm soil depth were found between PGH (58.8 ± 5.3 g kg−1 and 103 ± 7 g ha−1, respectively) and FDL (52.7 ± 4.3 g kg−1 and 117 ± 9 Mg ha−1, respectively) sampling. However, under Norway spruce, PGH sampling pointed‐out that C was mainly stored in Oe and Oa horizons (51.1 vs. 146 Mg ha−1 in the 0–30 cm layer). The higher coefficient of variation obtained when soil horizons were considered (from 19.2% to 72.8% for PGH and from 16.5% to 25.7% for FDL) suggested that PGH sampling preserved the information about the spatial variability within an ecosystem. The information loss about SOC vertical and spatial distribution would indicate the inability of FDL sampling to support decision‐making plans addressed for sustainable use of soil resource.