Abstract: Urban agglomerations are the main sources of carbon emissions. This paper utilized methods including carbon emission coefficient analysis, geometric centroid analysis, and multiple regression analysis to explore the spatiotemporal characteristics of land use carbon emissions in provincial urban agglomerations and their natural and socio-economic impact factors in Jiangsu Province. The results showed that: (1) Over 30 years, the total carbon emissions from land uses of the three agglomerations increased from 35.024 1 million tons to 328.75 million tons, spatially showing a clear gradient increasing distribution pattern from north to south. The total carbon emissions in Xuzhou urban agglomeration had grown faster than those in Nanjing and Su-Xi-Chang; The carbon emissions per unit area showed an overall upward trend, with the minimum and maximum values increasing from -3.99 and 214.7 t/hm2 to -2.74 and 262.43 t/hm2, respectively. (2) the carbon emissions per unit area in Nanjing and Su-Xi-Chang were significantly higher than those in Xuzhou urban agglomeration, but the growth rate in Xuzhou was faster; In the period of 30 years, the centroid of the total carbon emissions from land use changes had generally moved northward; (3) Factors such as topography and vegetation and ecological water bodies would have different degrees or directions of impact on the total and intensity of carbon emissions; (4) Among socio-economic factors, the total population was the most important factor driving the increase in total carbon emissions, but the impact was not the same in different urban agglomerations and at different stages of development; Population aggregation increased the scale of carbon emissions, but reduced the intensity of carbon emissions per unit area; The input of land elements in non-agricultural production, the structure of tertiary industry, and the increase in human activity intensity all promoted the increase in carbon emissions, and urbanization and socio-economic development were becoming increasingly important; The pulling effect of high-speed rail stations on economic activities was not been fully manifested. The proposed methodology was particularly suitable for the spatiotemporal differentiation of driving forces based on administrative units. This work provided support for the formulation of carbon reduction strategies for spatially differentiated provincial land uses, and thereby was helpful in achieving the "dual carbon" strategic goals.