Regional carbon cycle is the core content in global change research. Forest ecosystem in Da Xing'an mountain is sensitive to global climate change, thus, it is necessary to deeply understand the distribution of soil organic carbon density in this region. Here, four dominant forest types, i. e. larch forest, coniferous mixed forest, broadleaved mixed forest, and coniferous and broadleaved mixed fores were selected to investigate the variation of soil organic carbon density and plant community diversity in Da Xing'an Mountains of Northeast China by multi-factor analysis of variance. Results showed that Margalef, Shannon-Wiener, Mclntosh and Sorenson index showed larch forest>coniferous mixed forest>broadleaved mixed forest>coniferous and broadleaved mixed forest, while Cody index showed that coniferous and broadleaved mixed forest>broadleaved mixed forest> coniferous mixed forest>mixed forest. Soil organic carbon content and density showed the same change trend with the highest in the surface soil layer, and gradually decreased with the increasing of soil depth. Soil carbon density of coniferous and broadleaved mixed forest, broadleaved mixed forest, coniferous mixed forest, mixed forest was concentrated in the top 20 cm of soil which accounted for 35.24%, 31.61%, 31.70%, 32.39%, respectively. Correlation analysis showed that there was a significantly difference between Margalef, Shannon-Wiener, Mclntosh, Sorenson index and soil organic carbon content and density, while the correlation coefficient of soil organic carbon content was higher than soil organic carbon density. Double factor analysis showed that forest type had a large effect on soil organic carbon density (P<0.05), and forest type*soil depth had a large effect on soil organic carbon content (P<0.05). Forest type, forest type*soil depth had a large effect on Margalef and Shannon-Wiener index (P<0.05). In total, forest type and soil depth has the interaction effect on the soil organic carbon density.