This study identified material with high phosphorus sorption suitable for bioretention filter media. Materials examined were fly ash, two expanded shales, peat moss, limestone, and two common Oklahoma soils--Teller loam and Dougherty sand. The peat moss was a phosphorus source, while the two soils, limestone, and one expanded shale had only modest sorption capacity. One expanded shale and the fly ash had significant phosphorus sorption. Fly ash is unsuitable for use in a pure form, as a result of its low permeability, but phosphorus sorption on the sand was increased significantly with the incorporation of small amounts of fly ash. Column leaching experiments found that the sand with 2.5 and 5% fly ash and the better expanded shale had linear, non-equilibrium transport retardation factors of 272, 1618, and 185, with first-order rate coefficients of 0.153, 0.0752, and 0.113 hour(-1), respectively. Desorption experiments showed that the phosphorus sorption on the sand/fly ash mixture is largely nonreversible. Transport simulation assuming a 1-m-deep sand/fly ash treatment layer, with 5% of the watershed area, showed that the sand/fly ash filter media could effectively treat 1 mg/L influent for 12 years in a paved watershed and 34 years in a grassed watershed before exceeding Oklahoma's scenic rivers' phosphorus criterion of 0.037 mg/L. Significant phosphorus removal would continue for over 100 years.