Acorus gramineus Soland., Canna chineresisw, Calla palustris., Rhododendron simsii, Aspidistra elatior and Chamaedorea elegans, were rooted in 0.3m media. The media were composed of 0.15m gravel and 0.15m mixture of sand and soil, and the volume proportion of sand and soil was 2:1. The plants were grown in six 0.5x1.0m concrete ponds that were filled with a further 0.3m of effluent from an oxidation ditch operated in a sequential batch mode, treating eutrophic wastewater from a university. The water was sampled daily for total N and total P and retained for seven days. A control pond devoid of plants was not included. The levels of total P and total N declined to a maximum of 60-77.6% of initial for total P and 86.7-98% for total N within four days. Since levels stayed constant for total N but rose for total P, four day retention would minimize amounts of these nutrients leaving the ponds. The conclusion can be gotten through analysis that the plant’s absorb isn’t the main cause of total N and total P remove for constructed wetland, but there are some effect by the plant adaptability, so in the design process of eutrophic wastewater treatment, plants could be chosen on their aesthetic merits to enhance the wetland system’s sight effect. And the hydraulic residence is better inside 4 days in the surface constructed wetland design process. The mechanisms involved in nutrient decline were not investigated.