A better understanding of the discrepancies in multi-scale inventories could give an insight into their approaches and limitations as well as provide indications for further improvements; international, national, and plant-byplant data are primarily obtained to compile those inventories. In this study we develop a high-resolution inventory of Hg emissions at 0.05 • × 0.05 • for Jiangsu, China, using a bottom-up approach and then compare the results with available global/national inventories. With detailed information on individual sources and the updated emission factors from field measurements applied, the annual Hg emissions of anthropogenic origin in Jiangsu in 2010 are estimated at 39 105 kg, of which 51, 47, and 2 % were Hg 0 , Hg 2+ , and Hg p , respectively. This provincial inventory is thoroughly compared to three downscaled national inventories (NJU, THU, and BNU) and two global ones (AMAP/UNEP and EDGARv4.tox2). Attributed to varied methods and data sources, clear information gaps exist in multi-scale inventories, leading to differences in the emission levels, speciation, and spatial distributions of atmospheric Hg. The total emissions in the provincial inventory are 28, 7, 19, 22, and 70 % larger than NJU, THU, BNU, AMAP/UNEP, and EDGARv4.tox2, respectively. For major sectors, including power generation, cement, iron and steel, and other coal com-bustion, the Hg contents (HgC) in coals/raw materials, abatement rates of air pollution control devices (APCDs) and activity levels are identified as the crucial parameters responsible for the differences in estimated emissions between inventories. Regarding speciated emissions, a larger fraction of Hg 2+ is found in the provincial inventory than national and global inventories, resulting mainly from the results by the most recent domestic studies in which enhanced Hg 2+ were measured for cement and iron and steel plants. Inconsistent information on large power and industrial plants is the main source of differences in spatial distribution of emissions between the provincial and other inventories, particularly in southern and northwestern Jiangsu, where intensive coal combustion and industry are located. Quantified with Monte Carlo simulation, uncertainties in the provincial inventory are smaller than those of the NJU national inventory, resulting mainly from the more accurate activity data of individual plants and the reduced uncertainties in HgC in coals/raw materials.