[Objective] The influences of multi-level watershed forest changes on carbon stock changes were analyzed and the future trend of multi-level watershed carbon stock changes were predicted in order to provide a scientific basis for the sustainable development of regional forest land and the implementation of carbon neutral policy at the watershed scale. [Methods] The study was conducted for Chongqing City. We analyzed the spatiotemporal dynamic changes in forest land and carbon storage from 2000 to 2030 at the multi-level watershed scale by use of the PLUS model, landscape pattern indices, biomass expansion factor method, and gray correlation degree. Additionally, we determined the correlation between forest land pattern changes and carbon storage under natural development scenarios. [Results] ① Between 2000 and 2020, forest area increased by 3 404.55 km², primarily due to the conversion of cultivated land. The forest landscape pattern in the first-level watershed showed a tendency towards clustering and regularity, with a relatively reduced degree of fragmentation. There were distinct differences in the historical patterns and future trends of forest landscape pattern changes in the second-level watershed’s “district and group”. ② In the initial two decades, the carbon storage of forest land exhibited an increasing-decreasing trend, resulting in an overall increase. The Dongting Lake system in the first-level watershed experienced the highest increase in carbon storage, and the high-value area of carbon storage increment in the second-level watershed shifted from north to south. ③ Total carbon storage was predicted to increase by 2030, and the increase was strongly correlated with the irregular index. [Conclusion] The correlation between the forest landscape pattern index and carbon storage varied at different watershed scales. In the future, effective measures should be developed from a watershed perspective to improve carbon sink benefits and achieve carbon neutrality.