The composition of the microbial community in the rhizosphere was investigated using three different techniques: direct soil-phospholipid fatty acid (PLFA), cfu-PLFA (using culturable, bacteria only) and community-level physiological profiles (CLPP) using Biolog GN microtitre plates. All three techniques showed that the rhizosphere bacterial community differed from the bulk soil community, when pea plants were grown in pots. The two PLFA techniques also indicated that the effect of roots on microbial communities was greater than differences in the PLFA pattern between different bulk soils. The difference was most prominent using the direct soil-PLFA technique. Both PLFA techniques indicated that gram-negative bacteria were relatively more abundant in the rhizosphere, while PLFAs indicative of gram-positive bacteria were relatively more common in bulk soil samples. The results obtained with the direct soil-PLFA method did not differ between rhizosphere samples taken along a pea root, while this was the case with both the cfu-PLFA technique and the CUP method. According to both techniques, the differences in community composition were greatest near the root tip. Different plant species exhibited different microbial community composition in the rhizosphere, as judged from all three methods, when four different plant species were compared. In all cases, Trifolium differed most from the other species (three different grasses). Although all three methods were found to be suitable for rhizosphere studies, the CLPP method appeared less suitable than the two PLFA methods, since less of the variation in the data could usually be explained.