Membrane bioreactors (MBRs) were compared with conventional activated sludge systems (CAS) for micropollutant degradation, in laboratory-scale spiking experiments with synthetic and real domestic wastewater. The target micropollutants were polar in nature and represented a broad range in biodegradability. The experimental data indicated that MBR treatment could significantly enhance removal of the micropollutants 1,6- and 2,7-naphthalene disulfonate (NDSA) and benzothiazole-2-sulfonate. 1,5-NDSA, EDTA and diclofenac were not removed in either the MBR or the CAS. The other compounds were equally well degraded in both systems. For 1,3-naphthalene disulfonate, the existence of a minimum threshold level for degradation could be demonstrated. Although MBRs could not always make a difference in the overall removal efficiencies achieved, they showed reduced lag phases for degradation and a stronger memory effect, which implies that they may respond quicker to variable influent concentrations. Finally, micropollutant removal also turned out to be less sensitive to system operational variables.