It is often assumed that planted wastewater treatment systems outperform unplanted ones, mainly because plants stimulate belowground microbial population. Yet, fundamental interactions between plants and associated microorganisms remain only partly understood. The aim of our project was to evaluate microbial density and activity associated to the rhizosphere of three plant species. Experimental set-up, in six replicates, consisted of four 1.8-L microcosms respectively planted in monoculture of Typha angustifolia, Phragmites australis, Phalaris arundinacea and unplanted control. Plants were grown for two months with 25 L m(-2) d(-1) of secondary effluent (in g m(-2) d(-1): 1.3 TSS, 7.5 COD, 1.0 TKN). Sampling of substrate, roots and interstitial water was made according to depth (0-10, 10-20 cm). Biofilm was extracted with 500 mL of a buffer solution. Microbial density was directly estimated by flow cytometry and indirectly by protein measurements. Biological activity was determined using respirometry assays, dehydrogenase and enzymatic activity measurements. Our results show that microbial density and activity are higher in the presence of plants, with significantly higher values associated with Phalaris arundinacea. Greater density of aerobic or facultative bacteria was present in planted microcosm, particularly on root surface, suggesting root oxygen release. Microbes were present on substrate and roots as an attached biofilm and abundance was correlated to root surface throughout depth. Plant species root morphology and development seem to be a key factor influencing microbial-plant interaction.