Abstract: As the rapid development of the economy and agriculture, the water quality of many rivers and lakes is degrading due to agricultural non-point source pollution. The low polluted wastewater (LPW) is an eutrophic water body containing N, P and many microelements, meeting the Pollutant Discharge Standard of Municipal Wastewater Treatment Plant (Class one, B level). LPW can be reused before draining into downstream water system. Rural LPW mainly comes from agricultural production and farmer’s livelihood, including sewage tail water, farmland runoff and polluted river water. In this paper, the effect of C︰N ratio of LPW on the removal of nitrogen (N), phosphorus (P) and chemical oxygen demand (COD) of LPW was studied with floating-bed Oenanthe javanica system as the research subject to provide reference for increasing purification efficiency of floating-bed plant system. In the experiment, domestic sewage with or without glucose addition of floating-bed Oenanthe javanica system were two treatments (TWW-HC and TWW treatments), and no plant floating-bed system was used as the control treatment. The experiment lasted for 82 days with once water exchange in median time of the process. The results showed that N and P removal performance were better in TWW-HC treatment than that in TWW. External carbon addition rapidly reduced N and P contents, and the removal rates of total nitrogen (TN), ammonia nitrogen (NH⁺_4-N) and total phosphorus (TP) were 40.8%, 38.4%, and 62.8%, respectively, under TWW-HC treatment after 3 days of treatment. The removal rate of TN of TWW-HC was 73.9%96.0% during the whole experiment, which was higher than that of TWW (with the removal rate of 60.6%85.9%). The removal rate of total phosphorus (TP) was 68.0%81.1% in TWW-HC, which was higher than that in TWW and control treatments (with the removal rates of 21.3%54.9% for TWW and 19.2%58.1% for control, respectively). The biomass, average plant height, and relative growth rate of O. javanica were significantly higher in TWW-HC than in TWW. Plant uptake of TWW-HC accounted for 58.2% of the removed N amounts and 37.6% of the removed P amounts, greatly higher than those of TWW (8.7% of the removal N amounts and 11.0% of the removal P amounts). More than 80% of the removed N amounts were eliminated by other pathways in TWW treatment, but the percentage was only 37.3% in TWW-HC. Sedimentation contributed for 16.0% of the removed N amounts in control treatment, but only 8.5% in TWW and 4.6% in TWW-HC, respectively. Similar tendency was found in TP removal pathways. These results showed that the floating-bed O. javanica system evidently reduced N and P amounts in sediments compared to control. Results also indicated that increasing C︰N ratio to an appropriate range of LPW was beneficial for growth and nutrient uptake of O. javanica, which also contributed to N and P removal efficiently.