Prediction and making targeted measures to address global climate change requires precise estimate of local, regional, and global carbon budgets. Previously estimated global carbon budgets had significant uncertainty partly due to overlooking the soil carbon loss from soil erosion. However, as result of surface features being indistinguishable between eroded and non-eroded land, the lack of historical monitoring data, and confusion surrounding soil biochemical process, soil erosion was often overlooked by researchers. The overlooking of soil erosion may induce a great underestimate of soil carbon loss after land degradation or land use change. Through a case study in northeast China, we found that overlooking soil erosion underestimated 50% of soil carbon loss in degraded grassland. As a result, assessment of global carbon budget and climate trends in the past remained inaccurate. It is necessary to consider soil erosion as an important factor when calculating global carbon budget, and develop new approaches to monitor soil erosion process and track the intrinsic effects of soil erosion on soil carbon dynamics.