In 2023, the Australian Government issued ∼250,000 soil carbon credits following a measurement period characterised by high rainfall (Decile 10). The inferred soil organic carbon (SOC) sequestration rates during this period, ranging from ∼2 to 8 t C ha−¹ yr−¹, significantly exceed rates reported in Australian scientific studies (∼0.1 to 1.2 t C ha−¹ yr−¹). Our analysis, incorporating SOC and biomass measurements alongside remote sensing of NDVI, reveals that these SOC gains were largely attributable to above-average rainfall rather than project interventions. Moreover, these gains were not sustained when rainfall returned to average levels, raising concerns about the durability of credited sequestration and its additionality beyond natural climatic variability. Our findings demonstrate that current safeguards within the Soil Carbon Method—such as withholding 25% of credits during the first measurement period—are likely insufficient to account for climatic variability. To strengthen the integrity of the carbon crediting system, we recommend extending the minimum measurement period for credit issuance to at least five years. Additionally, governments should establish science-based ‘reasonable bounds’ for expected long-term SOC gains from management practices to sense-check reported outcomes. These measures will ensure that credited SOC sequestration is more closely tied to management-driven outcomes rather than short-term climate-driven fluctuations. A conceptual diagram of “new” carbon entering the soil system over a 25-year crediting period. Transient fluxes of SOC (blue) versus the accumulation of more persistent SOC (green). The risk of over crediting transient fluctuations of SOC is represented by the circle with a cross.