Precise spatiotemporal datasets of artificial impervious surfaces (AISs) are essential for evaluating urbanization processes and associated soil organic carbon (SOC) dynamics. However, spatially explicit studies on SOC stocks based on high-quality AIS data remain deficient, which affects the accuracy of urban SOC budgets. In this study, we used 30-m Landsat images and a subpixel-based model to accurately evaluate and quantify the annual AIS of Kaifeng, an ancient city in China that experienced intensive urbanization from 2000 to 2020. Soil organic carbon (SOC) dynamics were further estimated and spatially exhibited based on the SOC densities (SOCD) of different land covers observed in the field. Our results demonstrate that Kaifeng experienced drastic AIS expansion from 2000–2020, both in total area (an increase of ~154.35%) and density (described by mean AIS abundance, 0.56 vs. 0.72). Spatially, AIS mainly sprawled to the west, and infilling was observed in the old town. Moreover, the expansion of AIS in Kaifeng has resulted in a total of 0.08 Tg of SOC loss over the past 20 years, and the study area has acted as a clear carbon source. The greatest SOC losses occurred during 2010 — 2015, mainly in the west — with >30% (~0.024 Tg) of the total loss occurring between 2010 and 2015. This study provides new insights into urban growth through the mapping of growth patterns in terms of both outward sprawl and infill. We also provide a novel means of presenting the spatial patterns of urbanization-induced SOC dynamics using subpixel AIS maps.