The fate of selected UV filters (UVFs) was investigated in two soil aquifer treatment (SAT) systems, one supplemented with a reactive barrier containing clay and vegetable compost and the other as a traditional SAT reference system. We monitored benzophenone-3 (BP-3) and its transformation products (TPs), including benzophenone-1 (BP-1), 4,4'-dihydroxybenzophenone (4DHB), 4-hydroxybenzophenone (4HB), and 2,2'-dihydroxy-4-methoxybenzophenone (DHMB), along with benzophenone-4 (BP-4) and avobenzone (AVO) in all involved compartments (water, aquifer sediments, and biofilm). The reactive barrier, which enhances biochemical activity and biofilm development, improved the removal of all detected UVFs in water samples. Among monitored UVFs, only 4HB, BP-4, and AVO were detected in sediment and biofilm samples. But the overall retained amounts were several orders of magnitude larger than those dissolved. These amounts were quantitatively reproduced with a specifically developed simple analytical model that consists of a mobile compartment and an immobile compartment. Retention and degradation are restricted to the immobile water compartment, where biofilm absorption was simulated with well-known compound-specific Kow values. The fact that the model reproduced observations, including metabolites detected in the biofilm but not in the (mobile) water samples, supports its validity. The results imply that accumulation ensures significant biodegradation even if the degradation rates are very low and suggest that our experimental findings for UVFs and TPs can be extended to other hydrophobic compounds. Biofilms act as accumulators and biodegraders of hydrophobic compounds.