To separate the aggregates into the inner part (core) and the periphery (bark), the aggregates were rubbed against one another. For this purpose, aggregates of 5–3 mm were shaken 48 hours on a over head shaker, they rubbed against each other, and gradually their bark was grinded, passing into the fraction <0.25 mm. To monitor the process, the abrasion efficiency from the change in the size distribution of the size fractions was periodically evaluated. At the end of the ex-periment, the carbon and nitrogen contents were determined in the fractions obtained. The distribution of the resulting size fractions was stabilized after 16 hours of the experiment, in the case of Chernozem under the steppe, and after 24 hours in the case of Retisol under forest. In the Retisol abrasion (from the large aggregates, the peripheral part separates and particles <0.25 mm are formed) were observed throughout the entire experiment. In Chernozem, after the first hour of treatment, fragmentation of the aggregates predominated, then in the next 10 hours abrasion prevailed, then again crushing dominated. Determination of the carbon and nitrogen content shows the anisotropy of aggregates in this parameter. Inside the aggregates, the carbon content is higher than at the periphery, with the C/N ratio indicating the predominance of poorly-decom-posed plant residues. Moreover, these differences are more expressed in uncultivated soils, and when comparing types, in Retisols in comparison with Chernozems. In general, the proposed approach is promising. However, it requires further development for an effective application.