For advanced phosphorus (P) removal sorption processes were studied in benchscale to offer a post-treatment step in addition to the relatively unknown biological Premoval in membrane bioreactors (MBR) for decentralized wastewater treatment. First three commercially available sorbents, granulated ferric hydroxide GFH, activated alumina ATE and an iron hydroxide/oxide FER, were screened for their affinity and capacity in removing phosphorus from MBR filtrates. GFH features the highest loading and affinity for P and additionally removes organics followed by ATE. For example, in determination of isotherms at a P-equilibrium concentration of 0.5 mg/L the loading on GFH amounts to 20 mg/g, on ATE to 9 mg/g and on FER only to 3.8 mg/g. Very early FER was excluded from the trials due to its low capacity and desorption of bonding agents leading to an increase instead of a removal of dissolved organic carbon. Different influencing parameters were studied in the P-sorption on GFH and ATE. Lowering the pH from 8 to 5 improved the P-sorption only by 30 %. A pH-adjustment for optimization is not recommendable due to the strong buffering of MBR filtrates by the carbonate/hydrogen carbonate system. A decrease in temperature had no effect in P-sorption on GFH at low initial P-concentrations, while the loading on ATE was reduced compared to room temperature. Therefore, an improvement is not expected by an outdoor operation of sorption filters. No main influencing competition by other organic or inorganic compounds could be determined on P-sorption in MBR filtrates. Due to the negatively charged sorbent surfaces the specifically sorbing phosphorus has an decisive advantage over unspecifically sorbed anions like chloride and sulfate. In time depending experiments lacking of calcium had a kinetic effect on P-sorption, but calcium is present in sufficient amount for P-sorption in MBR filtrates. In sorption filter tests using bed volumes (BV) around 90 mL a limit value of 50 µg/L P is reached in the effluent at 8,000 bed volumes for GFH and 4,000 for ATE sing a influent concentration of 0.3 mg/L P. At a lower influent concentration of 0.1 mg/L P the breakthrough at the limit can be extended to above 15,000 BV for GFH and 8,000 for ATE. While GFH affords a contact time of 10 min, around 30 min contact time is needed in sorption filters filled with ATE to achieve a late breakthrough. A partial regeneration and P-reloading on GFH and ATE can be achieved by using 0.6 M NaOH in differential recirculating flow reactors. But for several reasons the regeneration can not be transferred to sorption filters directly. Due to the long operation time of sorption filters at low influent concentrations = 0.3 mg/L P the regeneration and reloading could not be studied in sorption filters here. Therefore, the cost estimations could only be based on a few scenarios. Using ATE or GFH, total costs are in the range of 8 - 30 Cents/m³ depending on the possibility of regeneration and reuse of sorbents. If no regeneration can be achieved, an application of ATE would be favoured due to its lower price. Based on an effective regeneration a sorption technique using GFH would be preferable. This is emphasized by the high affinity of GFH for P, which is especially required at low effluent P-concentrations from MBR. As an outlook it is recommended to verify the sorption performance and to study the breakthrough behavior in a pilot scale (H = 1 m, d = 300 mm). Furthermore, the investigations on regeneration have to be intensified and conducted in sorption filters to assess influencing parameters like the effect of the previous sorption time on regeneration, the regeneration time and concentration of regenerate. The frequency of regeneration and reuse of sorbent until disposal can only be stated by future longterm trials in sorption filters. Due to the low volume flow rate in the small studied sorption filter, which showed no pressure drop during the trials, it is hard to predict if a backwashing is necessary in larger scale. Whether the particle-free MBR filtrate is advantageous for filter operation, in a long-term scale microorganism growth might occur and result in clogging of sorption filters. In this view a regeneration seems to be advantageous and might replace both backwashing and disinfection rinsing.