The construction of artificial wetlands has become a measure increasingly applied to reduce nonpoint-source (NPS) pollution and to contribute to the restoration of eutrophic lakes and coastal waters. In a 2-yr study monitoring fluxes of particulate and dissolved phosphorus (P) in a small artificial wetland for the treatment of agricultural drainage water in Central Switzerland, water residence time was identified as the main factor controlling P retention in the system. Since most of the annual P load (62% as dissolved reactive phosphorus, DRP) was related to high discharge events, it was not average but minimum water residence time during flood events that determined the wetland's P retention. In agreement with a continuous stirred tank reactor (CSTR) model, our investigations suggest a minimum water residence time of 7 d to retain at least 50% of the bioavailable P. The investigated wetland retained only 2% of the bioavailable P, since the water residence time was shorter than 7 d during 61% of time in both years. Settling of phytoplankton rather than DRP uptake into phytoplankton limited the retention of bioavailable P. The overall retention efficiency of 23% total phosphorus (TP), corresponding to a surface related retention of 1.1 g P m(-2) yr(-1), was due to the efficient trapping of pedogenic particles.