The aim of the study was to examine the differences in microbial community structure as a result of agricultural practices. Sixteen samples of cultivated and the same number of non-cultivated soils were selected. Gel bands were identified using the GelCompar software to create the presence-absence matrix, where each band represented a bacterial operational taxonomic unit. The data were used for principal-component analysis and additionally, the Shannon-Weaver index of general diversity, Simpson index of dominance and Simpson index of diversity were calculated. Denaturing gradient gel electrophoresis profiles clearly indicated differentiation of tested samples into two clusters: cultivated and non-cultivated soils. Greater numbers of dominant operational taxonomic units (65) in non-cultivated soils were noted compared to cultivated soils (47 operational taxonomic units). This implies that there was a reduction of dominant bacterial operational taxonomic units by nearly 30% in cultivated soils. Simpson dominance index expressing the number of species weighted by their abundance amounted to 1.22 in cultivated soils, whereas a 3-fold higher value (3.38) was observed in non-cultivated soils. Land-use practices seemed to be a important factors affected on biodiversity, because more than soil type determined the clustering into groups.