Due to anthropogenic practices, the sludge of urban origin is contaminated by organic pollutants such as polycyclic aromatic hydrocarbons (PAH) and polychlorobiphenyls (PCB). Efficient biodegradations under mesophilic methanogenic conditions were already reported for these micropollutants (Trably et al 2003; Barret et al 2010). Here, we have determined the influence of the microbial community on the micropollutants removal while physico-chemical conditions are strictly controlled. We investigated three microbial communities extracted from ecosystems with contrasting pollution histories; PAH contaminated soil, PCB contaminated sediment and moderately contaminated anaerobic sludge. We used an original combination of enzymatic treatments and a cell flotation step to harbour the communities free of their surroundings (Braun et al 2011). Each microbial community served as inoculum for three 400mL anaerobic digesters. A total of 12 mesophilic continuous reactors were operated during 100days with a hydraulic retention time (HRT) of 20days. The substrate was a sterilised activated sludge, spiked with 13 PAH and 7 PCB (5mg.kg-1 DM). The organic matter degradation, the biogas production rate and composition, and the micropollutants removals were monitored. The bacterial and archaeal communities were characterized in abundance (qPCR) and community structure (SSCP fingerprinting). Population shifts were quantified using diversity index and principal component analysis (PCA). Functional steady states, based on organic matter degradation, were observed after 3-4 HRT. The PAH removals varied from 5 to 30% depending on the molecules but not on the inoculum, and were highly correlated with the organic matter removal (cor=0.4475-0.614, p-value<0.001). By contrast, the dynamics of biogas productions and the biogas compositions differed according the inoculums tested, with increasing CH4/CO2 ratios for PAH soil, PCB sediment, and anaerobic sludge. Greater PCB removal were observed for the reactors inoculated with PCB contaminated sediment. The bacterial densities reached values of 1012 Bacteria.gVS-1 at steady state for all reactors. By contrast, the density of Archaea varied from 3.8x109 to 3.5x1010 Archaea.gVS-1 depending on the origin of inoculum (PAH soil < PCB soil < anaerobic sludge). The bacterial and archaeal genetic structures were compared with fingerprinting data using PCA. The inoculum significantly influenced the genetic structure of bacterial (cor=0.321, p-value<0.001) and archaeal (cor=0.420, p-value<0.001) communities. The diversity level was also affected by the inoculum used to seed anaerobic reactors (cor=0.774 for Bacteria and cor=0.310 for Archaea, p-value=0.001). The reactors inoculated with the anaerobic sludge displayed always higher diversities than the others. A diversity-functioning relationship was evidenced, when comparing diversity indices and VFA concentrations (cor=0.669 for Bacteria and cor=0.537 for Archaea, p-value=0.001), but no correlation between microbial communities and micropollutants removal were observed. As a conclusion, we obtained three functional stable consortia with contrasted macroscopic functioning and microbial dynamics, but similar pollutant removals. The removal of persistent organic pollutants did depend on the molecule type, but did not depend on the metabolic functioning during methanogenesis. The bioavailability of micropollutants in regard with their biodegradation will be discussed.