A field experiment was conducted to remove the potential eutrophication effect of P from rural wastewater (RW) during the whole rice growing season of 2007. The experiments consisted of five treatments, namely black water (BW). domestic wastewater (DW), grey water (GW), surface lake water (SW) and surface lake water without P application as a check (CK), with three replicates in a randomized block design. Commercial fertilizer and RW were applied to furnish 40 kg P ha(-1) except CK. Results showed total P (TP) concentration had significantly declined after P application, from October 15 there were no significant increases in TP concentration in the floodwater. TP removal rates from RW was significantly higher(P <= 0.05) than those from fertilizer. TP load was in an overall gradual decline, whereupon it became approximately steady on October 1. The percentage of TP load from wastewater decreased, whereas that from fertilizer continued to increase. Meanwhile, the yield for CK was significantly less (P <= 0.05) than SW, GW, DW, and BW, with the yield of BW significantly greater (P <= 0.05) than other treatments. It is feasible to remove P from RW by the paddy-rice-wetland system and can be widely used to improve the yield of rice. Crown Copyright (C) 2009 Published by Elsevier B.V. All rights reserved.