Agricultural soils have lost vast amounts of carbon since the plow was invented — which means they have vast capacity to take it back. Soil carbon credits pay farmers to rebuild that stock through regenerative practice. It’s one of the most promising and most methodologically contested corners of the carbon market.
How farming stores carbon
Plants pump carbon into soil through roots and residues; microbes and minerals stabilize a fraction of it as soil organic carbon. Conventional intensive agriculture — deep tillage, bare fallows, heavy synthetic inputs — burns that stock down. Regenerative practices reverse the flow: no-till or reduced tillage keeps carbon in the ground; cover crops keep living roots feeding the soil year-round; diverse rotations, composting, and managed grazing build organic matter. Healthy soils also hold more water and often cut input costs, which is why many farmers adopt these practices credits or not.
How the credits work
A project enrolls farmland, documents a practice change (the additionality test — business-as-usual farming earns nothing), and quantifies the resulting carbon gain against a baseline. Credits are issued per verified tonne of CO2e sequestered. Programs from Verra, Gold Standard, and agricultural platforms differ in one crucial respect: how they measure.
The measurement debate
Soil carbon changes are small (fractions of a percent per year), slow, and patchy across a field — detecting them reliably requires either dense, expensive soil sampling or models calibrated on sparse samples. Model-heavy programs issue credits cheaply but attract skepticism; sampling-heavy programs are rigorous but costly. The frontier is hybrid MRV: spectroscopy, remote sensing, and process models anchored by physical cores. Buyers should know which approach backs their credits — it’s the single biggest quality variable.
Permanence and price
Soil carbon is reversible — one deep tillage pass can release years of gains — so programs use long-term contracts, buffer pools, and monitoring to manage reversal risk. Prices in 2026 typically run $15–40 per tonne depending on rigor and region: above forestry avoidance, below biochar, reflecting exactly that mix of removal status and permanence uncertainty.
Why it still matters enormously
Agriculture occupies nearly 40% of Earth’s land. Even modest per-hectare sequestration, multiplied across that area, is gigatonne-scale potential — with food security, drought resilience, and farmer income riding along. For buyers, rated soil projects with strong sampling regimes make a defensible portfolio slice. For agriculture, carbon may become what it’s really about: a second crop grown underground.
Related reading: Biochar Carbon Credits · Carbon Credit Additionality Explained
