We show that acid type did not significantly affect results. We also find that (i) glacial and interglacial samples do not react similarly to acid pre-treatment, (ii) high acid strength (>1.5 mol L-1) induces significant bias on δ13Corg and TOC values, and therefore on C/N ratios; (iii) 25% of an isotopically distinct pool of organic carbon was lost when using acid concentrations of 0.2-1.5 mol L-1, affecting δ13Corg values by more than 1.5‰; and (iv) geochemical evidence indicates that the leachable organic carbon pool is preferentially composed of terrestrial organic matter. These findings call for precautions when using C/N ratios and associated δ13Corg values for paleoenvironmental and climate reconstructions.