Rhizosphere bacterial communities of two transgenic potato lines which produce T4 lysozyme for protection against bacterial infections were analyzed in comparison to communities of wild-type plants and transgenic controls not harboring the lysozyme gene. Rhizosphere samples were taken from young, flowering, and senescent plants at two field sites in three consecutive years. The communities were characterized in a polyphasic approach. Cultivation-dependent methods included heterotrophic plate counts, determination of species composition and diversity based on fatty acid analysis of isolates, and community level catabolic profiling. Cultivation-independent analyses were based on denaturing gradient gel electrophoresis (DGGE) of 16S rRNA gene fragments amplified from rhizosphere DNA using primers specific for Bacteria, Actinomycetales, or alpha- or beta-Proteobacteria. Several bands of the DGGE patterns were further characterized by sequence analysis. All methods revealed that environmental factors related to season, field site, or year but not to the T4 lysozyme expression of the transgenic plants influenced the rhizosphere communities. For one of the T4 lysozyme-producing cultivars, no deviation in the rhizosphere communities compared to the control lines was observed. For the other, differences were detected at some of the samplings between the rhizosphere community structure and those of one or all other cultivars which were not attributable to T4 lysozyme production but most likely to differences observed in the growth characteristics of this cultivar.