Abstract: A long-term thinning experiment in Korean larch (Larix olgensis) plantation of Mengjiagang Forest Farm of Kiamusze in hilly area of Sanjiang Plain, northeastern China, was conducted to identify the effects of thinning on biomass carbon (dead and live biomass), soil carbon (forest floor and mineral soil), total ecosystem carbon storage and accumulative carbon sequestration (dead tree +thinned tree+ ecosystem carbon storage). Two harvesting scenarios (stem-only vs. whole-tree harvesting) were assessed in terms of carbon export. The study site was a 56-year-old larch plantation, where five thinnings of different intensities and frequencies were applied: 2 times heavy (35.6%-43.4%) thinning (L1), 2 times moderate (23.1%-24.3%) thinning (L2), 3 times moderate (15.3%-23.8%) thinning (L3), 4 times light (5.8%-17.1%) thinning (L4) and unthinned (CK, only harvesting dead biomass). The five thinning interventions involved whole-tree harvesting of thinned and dead trees (entire removal of slash and stem). The results revealed difference between the unthinned and thinned plots as regards the total dead wood debris, the former containing 40.3 t/ha, in the case of thinned plots, 8.3 t/ha (range 3.1-14.1). The energy wood (logging residues) and timber production by thinning were 10.4 t/ha (range 6.9-13.8) and 32.8 m3/ha(rang 21.9-50.1), respectively. Although the forest floor carbon pool was susceptible to loss (14.8% lower than CK), the mineral soil carbon pool could be enhanced by thinning (5.6% higher than CK), particularly in L3 plot (15.5% higher than CK). Thinning could not change the allometric relationships of living tree biomass carbon pool (proportions of wood-stem, root, branch, bark and foliage to the retained living tree biomass carbon pool were 67.7%-68.7%, 17.5%-18.0%, 6.8%-7.0%, 4.8%-4.9% and 2.2%-2.3%, respectively) and total ecosystem carbon pool (proportions of retained living trees, 0-30 cm mineral soil, forest floor, dead standing tree, shrub layer and herb layer to the total ecosystem carbon storage were 69.7%-72.0%, 24.7%-27.7%, 1.5%-2.2%, 0-1.3%, 0.1%-1.3% and 0.1%-0.2%, respectively), but increase the proportion of belowground carbon (roots of live and standing dead trees+0-30cm mineral soil+forest floor+shrub+herb) to the total ecosystem carbon storage (40.5%-42.4%, 40.0% for CK), and decrease the proportion of aboveground carbon (stem+branch+bark; 56.0%-57.9%, 58.3% for CK). Stem-only harvesting of old growth plantation could leave 26.5%-27.4% of living biomass carbon (27.7% for CK), whole-tree harvesting could leave19.7%-20.3% of living biomass carbon (20.5% for CK), and 42.1%-44.0% of ecosystem carbon (41.7% for CK). The averaged values of retained tree biomass, ecosystem carbon storage and stand accumulative carbon sequestration of thinned old growth larch plantations (56-year-old) were of similar levels with CK, and only 1.7% (-4.3%-1.5%), 1.7%(-5.9%-1.4%) and 1.1%(-4.0%-0.8%) lower than CK, respectively, but in the L3 and L4 plots, particularly in the L4 plot, the above indexes were 1.5%, 1.4%, and 0.8% higher than CK, respectively. Our results indicated that 3 and 4 times light or moderate (5.8%-23.8%) thinning not only supplies energy wood and timber production, but also does not alter the retained tree biomass, total ecosystem carbon content and stand accumulative carbon sequestration of old growth larch plantation, suggesting the sustainability of these silvicultural treatments.