Two shock loads of a commercial detergent (I-150 mg chemical oxygen demand (COD)/L, fed for 56 h; II-300 mg COD/L fed for 222 h) were applied in a lab-scale Expanded Granular Sludge Blanket (EGSB) reactor, fed with 1,500 mg COD/L of ethanol. The impact of the surfactant was assessed in terms of granular sludge morphology, specific methanogenic activity (SMA) in the presence of individual substrates, and reactor performance. COD removal efficiency remained unaffected in the shock 1, but 80 h after starting exposure to the shock II, the COD removal efficiency decreased drastically from 75 to 17%. In the first 8 h of operation of shock 1, the SMA was stimulated and decreased afterwards, being recovered 5 days after the end of exposure time. Concerning to shock II, the SMA was immediately and persistently reduced during the exposure time, although, the inhibition of SMA in presence of H-2/ CO2 showed a trend to increase after the exposure time. Acetoclastic bacteria were observed as the most sensitive to the toxic effects of surfactant whereas the hydrogenotrophic bacteria were less affected. The inhibitory effects were dependent on surfactant concentration and exposure time. The ratio filaments length per total aggregates area (LfA) was an early-warning indicator of biomass washout, since it increased 3 and 5 days before effluent volatile suspended solids (VSS) rise, respectively, in shocks I and II.