(1) A study was made of the effect of secondarily treated effluent on the phosphorus budget of a central Florida freshwater marsh. The marsh was chiefly composed of Sagittaria lancifolia, Pontederia cordata, Panicum spp. and Hibiscus sp. Three 2000 m² plots received effluent at the rates of 1·3, 3·8 and 10·2 cm wk-1, while a fourth 2000 m² plot was a control plot which received 3·8 cm wk-1 of freshwater. (2) In the first year the plot receiving the high rote of effluent showed increased net production of plant shoots, increased litter production, increased root and rhizome production and higher phosphorus concentrations in living and dead plant tissue compared with the control plot. (3) Effluent treatments did not significantly increase the phosphorus concentrations measured in the groundwater draining from the experimental plots. All the wells within the effluent plots had phosphorus concentrations c. 97% less than the phosphorus concentration of the applied effluent. (4) Phosphorus budgets were constructed for the control plot and the plot receiving the high effluent treatment. Over the course of the study 38·03 g P m-2 were applied to the high effluent plot. Of the total input 26·31, 8·81 and 1·97 g P m-2 were stored in the soil, roots and rhizomes, and litter respectively. Outflow of phosphorus from the plot amounted to only 0·94 g P m-2 The freshwater control plot received a total of 0·38 g P m-2. Storage in the litter accounted for 0·21 g P m-2 and outflow accounted for the remaining 0·17 g P m-2. There was no evidence of a spring flush of phosphorus from either the control or effluent plot. (5) The marsh successfully removed phosphorus from the effluent during the first year of application. It was concluded that long-term use of the marsh for phosphorus removal may be contingent upon the phosphorus adsorption capacity of the soil and the rate of peat production. [ABSTRACT FROM AUTHOR]