Shifting cultivation is a predominant land use across the tropics, feeding hundreds of millions of marginalised people, causing significant deforestation, and encompassing a combined area of land ten-fold greater than that used for oil palm and rubber. A key question is whether carbon-based payment for ecosystem services (PES) schemes can cost-effectively bring novel restoration and carbon-sensitive management practices to shifting agriculture. Using economic models that uniquely consider the substantial area of fallow land needed to support a single cultivated plot, we calculated the break-even carbon prices required for PES to match the opportunity cost of intervention in shifting agriculture. We do so in the North-east Indian biodiversity hotspot, where 35.4% of land is managed under shifting agriculture. We found net revenues of US$829.53-2581.95 per 30 ha when fallow area is included, which are an order of magnitude lower than previous estimates. Abandoning shifting agriculture entirely is highly feasible with break-even prices as low as US$1.33 t −1 CO 2 , but may conflict with food security. The oldest fallow plots could be fully restored for US$0.89 t −1 CO 2 and the expansion of shifting agriculture into primary forest halted for US$0.51 t −1 CO 2 , whereas abandoning short-fallow systems would cost US$12.60 t −1 CO 2 . A precautionary reanalysis accounting for extreme economic uncertainty and leakage costs suggests that all interventions, excluding abandoning short-fallow systems, remain economically viable with prices less than US$4.00 t −1 CO 2 . Even with poorly formed voluntary carbon markets, shifting agriculture represents a critical opportunity for low-cost forest restoration whilst diversifying income streams of marginalised communities across a vast area.