Groundwater is an important drinking water resource. Yet, this resource is threatened by pollution from chemicals, such as pesticides and their degradation products. To investigate the potential for remediation of groundwater polluted by trace concentrations of the pesticide residue 2,6-dichlorobenzamide (BAM), we established a pilot waterworks including two sand filters. The waterworks treated groundwater polluted with 0.2 μg/L BAM at flow conditions typical for rapid sand filters. Bioaugmentation of the sand filter with a specific BAM-degrading bacterium (Aminobacter sp. MSH1) resulted in significant BAM degradation to concentrations below the legal threshold level (0.1 μg/L), and this without adverse effects on other sand filter processes such as ammonium and iron oxidation. However, efficient degradation for more than 2-3 weeks was difficult to maintain due to loss of MSH1-bacteria, especially during backwashing. By limiting backwash procedures, the period of degradation was prolonged, but bacteria (and hence degradation activity) were still lost with time. Protozoa were observed to grow in the filters to a density that contributed significantly to the general loss of bacteria from the filters. Additionally, the concentration of easily assimilable organic carbon (AOC) in the remediated water may have been too low to sustain a sufficient population of degrader bacteria in the filter. This study shows that scaling up is not trivial and shortcomings in transferring degradation rates obtained in batch experiments to a rapid sand filter system are discussed. Further optimization is necessary to obtain and control more temporally stable systems for water purification. However, for the first time outside the laboratory and at realistic conditions a potential for the biodegradation of recalcitrant micropollutants in bioaugmented rapid sand filters is shown.