Soil is an important reservoir of biogenic carbon (C) at the ecosystem and global scales. This reservoir is characterized by thermodynamic stability of organic compounds and high activity in the modern environment. It is now well known that the soil C pool is not homogeneous, but is comprised of different chemical and physical fractions which differ in stability. In this study, we quantified the physical and chemical characteristics of soil organic matter in a Russian Chernozem under different land cover and land uses. The study site was located in the Central Chernozem Region of Russia (51003' N, 40042' E). Soils were collected from 3 adjacent ecosystems: virgin steppe (never tilled, but cut annually to prevent afforestation), a 102 yr old forest, and an agricultural field cropped as a corn monoculture for the past 45 yrs. We used particle size-density fractionation to separate roots, 3 light or free organic fractions (<1.8 g/cm3 <45 µm, <1.8 g/cm3 >45 µm, and 1.8-2.0 g/cm3) and 4 organo-mineral fractions (<0.2 µm, 2-0.2 µm, >2 g/cm3 <2 µm and >2 g/cm3 >2 µm) using ultrasonic treatment, sedimentation, and Na-polytungstate solutions for density separations. Organic matter in each fraction was analyzed by isotope ratio mass spectrometry, hydrophobic interaction chromatography, IR-spectroscopy, and traditional chemical methods. Compared to the native steppe, soil organic C increased under forest and decreased in cultivated soil; the largest ratio of humic/fulvic acids (HA/FA) was found in agricultural soil, while steppe and forest soils had similar ratios. In all 3 ecosystems, approximately 75% of the soil was present in only two of the organo-mineral fractions - heavy coarse (>2 g/cm3 >2 µm), and coarse clay (0.2-2 µm), containing about 0.3 and 7.0% C, respectively. Light fractions of free organic matter (<1.8 g/cm3 < and >45 µm, 1.8-2.0 g/cm3) were relatively minor components of this soil (up to 15, 4 and 8 %, respectively), but contained up to 25, 20 and 10 % organic C, respectively, and stored more than half the total stock of organic C. The forest soil was relatively enriched in the light fractions, whereas the agricultural soil was relatively depleted in these fractions. The light fraction ranged from 25% C in the agricultural soil to 18% C in the forest soil. Thus, both the absolute mass and organic C concentration of the light free organic matter fractions were the most sensitive to land cover and land use change. Previously, we showed that ä15N of soil physical fractions had a strong positive correlation with the mean residence time (MRT) of the organic matter in those fractions. In this study, ä15N of soil fractions increased in the following order: (1) roots, (2) <1.8 g/cm3 and >45 µm, (3) <1.8 g/cm3 and <45 µm, (4) 1.8-2.0 g/cm3, (5) 2-0.2 µm, (6) >2 g/cm3 and >2 µm, (7) <0.2 µm, (8) >2 g/cm3 and <2 µm. This implies that organic matter in the smaller, heavier fractions has the longest MRTs. C/N ratios in above order of separates demonstrate a gradual decrease from 30 to 7 in all 3 ecosystem types. The same order of fractions also showed nonrandom regular variations of ä13C that were similar for all soils. Roots had ä13C = –26.4 ± 1‰ in steppe and forest, and –-22.1‰ for corn field. ä13C then decreased from roots to the <1.8 g/cm3 >45 µm fraction (~1‰ less for the steppe and forest soil, ~4‰ less for soil under corn), then gradually increased to the >2 g/cm3 <2 µm fraction (~2‰ greater for all soils). For steppe and forest soils, the changes of ä13C were in close agreement numerically. So, two stages could be distinguished – depletion in 13C produced by decay of plant material, and enrichment in 13C connected with humification. Chromatography of organic matter associated with clay minerals indicated comparatively high proportions of hydrophylic organic components representative of fulvic acids. This was true for all 3 soils. Free organic matter contained hydrophobic substances more typical of humic acids. The highest proportion of hydrophobic components was detected in agricultural soil. These chromatographic results related to the distribution of humic and fulvic acids were further confirmed by IR-spectroscopy. Based on our methods, the soil organic C pool appears to be comprised of 3 main components: (1) Plant residues with a very high rate of turnover; (2) Thermodynamically stable, high rate of turnover, relatively high content of hydrophobic, aromatic free organic compounds; and (3) Thermodynamically stable, relatively low rate of turnover, more hydrophilic and aliphatic organic compounds associated with mineral particles. The first two constituents are capable of rapid response to land cover/land use changes, while the third constituent is more resistant to these changes. Total C for this chernozem is distributed between constituents in the approximate ratio 1:10:9.