Future stringent phosphorus regulations (down to 50 microg/L in some cases) together with the availability of more cost effective and/or innovative membrane processes, are the bases for this project. In contrast to conventional activated sludge plants, process parameters are not optimised and especially enhanced biological phosphorus (Bio-P) removal in membrane bioreactors (MBRs) are not proven yet. Current practice of P-removal in MBRs is the addition of coagulants in a co-precipitation mode. Enhanced biological phosphorus removal, when adapted to MBR technology, might be a cost-effective process. For very stringent effluent criteria additional P-adsorption on activated clay after membrane filtration can be also an interesting solution. The objective of this research project is to identify and test various phosphorus removal processes or process combinations, including MBR technologies. This should enable us to establish efficient and cost effective P-removal strategies for upgrading small sewage treatment units (up to 10,000 PE), as needed in some decentralised areas of Berlin. In particular, enhanced Bio-P removal technology was developed and optimised in MBR. Combinations of co-precipitation and post-adsorption will be tested when low P-values down to 50 microg/L are required in the effluent. One MBR bench-scale plant of 200 to 250 L and two MBR pilot plants of 1 to 3 m3 each were operated in parallel to a conventional wastewater treatment plant (Ruhleben WWTP, Berlin, Germany). The MBR bench-scale and pilot plants were operated under sludge ages of respectively 15 and 25 days. In both cases, Bio-P was possible, and phosphorus effluent concentration of about 0.1 mg/L could be achieved. A similar effluent quality was observed with the conventional WWTP. Investigations with lab columns indicated that P-adsorption could lead to concentrations down to 50 microg/L and no particle accumulation occurred in the filter media. The three tested materials exhibited great differences in break-through curves. Granulated ferric hydroxyde (GEH) showed higher capacity than activated alumina and FerroSorpPlus.