Reducing greenhouse gas emissions is one of the major challenges in combating global warming. Carbon, including in the form of carbon dioxide (CO₂), is considered an essential greenhouse gas under human control to demonstrate success in emission reductions. However, many carbon stock quantifications in forest ecosystems still rely on the estimated 50% carbon content instead of more precise species-, tissue- and site-specific values. Thus, this study aimed to thoroughly measure and analyze the carbon content and variability using the 14 major tree species in Northeast China. Over 600 trees were destructively sampled from three different major mountainous regions (i.e., the Changbai, Daxing’an, and Xiaoxing’an mountains), and the carbon contents of each species were precisely measured to the sub-tissue level. Carbon contents varied significantly between species, with foliage carbon mostly found to be the highest, while root carbon contents were the lowest. Average carbon contents can be ranked as: Ulmus laciniata (43.4%) < Phellodendron amurense (43.5%) < Acer mono (43.8%) < Tilia amurensis (44.2%) < Populus davidiana (44.5%) < Fraxinus mandshurica (44.7%) < Juglans mandshurica (44.9%) < Quercus mongolica (45.3%) < Betulla davurica (45.8%) < Betulla platyphylla (46.7%) < Picea koreansis (46.9%) < Larix gmelinii (47.4%) < Pinus koreansis (48.3%) < Abies nephrolepis (48.3%). Carbon contents were higher in conifers (47.7%) compared to broadleaf species (44.9%). In addition, both tree tissues and growing sites also had a significant effect on carbon content. At the sub-tissue level, only stem’s sub-tissues (i.e., bark, heartwood, and sapwood) carbon contents showed significant variations. The results suggest that bark should be separated from other stem sub-tissues and considered separately when determining carbon stocks. This research contributes to improving estimates of terrestrial carbon quantifications, and in particular, the values obtained can be used in China’s National Forest Inventory.