A fluidized bed bioreactor (FBBR) was operated for more than 575 days to remove 2,4,6-trichlorophenol (TCP) and phenol (Phe) from a synthetic toxic wastewater containing 80 mg L-1 of TCP and 20 mg L-1 of Phe under two regimes: Methanogenic (M) and Partially-Aerated Methanogenic (PAM). The mesophilic, laboratory-scale FBBR consisted of a glass column (3 L capacity) loaded with 1 L of 1 mm diameter granular activated carbon colonized by an anaerobic consortium. Sucrose (I g COD L-1) was used as co-substrate in the two conditions. The hydraulic residence time was kept constant at I day. Both conditions showed similar TCP and Phe removal (99.9 + %); nevertheless, in the Methanogenic regime, the accumulation of 4-chlorophenol (4CP) up to 16mgL(-1) and phenol up to 4mgL(-1) was observed, whereas in PAM conditions 4CP and other intermediates were not detected. The specific methanogenic activity of biomass decreased from 1.01 +/- 0.14 in M conditions to 0.19 +/- 0.06 mmol CH4 h(-1)gTKN(-1) in PAM conditions whereas the specific oxygen uptake rate increased from 0.039 +/- 0.008 in M conditions to 0.054 +/-0.012 mmolO(2) h(-1) gTKN(-1), which suggested the co-existence of both methanogenic archaea and aerobic bacteria in the undefined consortium. The advantage of the PAM condition over the M regime is that it provides for the thorough removal of less-substituted chlorophenols produced by the reductive dehalogenation of TCP rather than the removal of the parent compound itself. (c) 2005 Society of Chemical Industry.