Variations in phosphorus speciation in two sets of simulated riparian zones with and without Perennial ryegrass were compared. Each set consisted of four units, each measuring 700 mm × 200 mm × 200 mm, which were enhanced with 0, 2.5, 5, and 7.5% red mud (RM) by weight. The levels of total phosphorus (TP), total dissolved phosphorus (TDP), soluble reactive phosphorus (SRP) in the effluent were analyzed, and phosphorus fractionation in the media were also determined after the systems had been operational for 3 months. The results showed that the unit received 2.5% RM had the highest rate of phosphorus removal, including TP, TDP, SRP, particulate phosphorus (PP), and dissolved organic phosphorus (DOP) were present at the average concentrations of 0.17, 0.10, 0.07, 0.08, and 0.03 mg/L in the effluent. Sequential phosphorus fractionation showed that calcium-bound phosphorus (Ca–P) was the major component, indicating that the addition of RM induced aluminum/iron-bound phosphorus (Al/Fe–P), which was intensely bioactive, to form intractable Ca–P, which further inhibited the release of phosphorus from the media. However, the presence of P. ryegrass had little effect on the removal of phosphorus. Therefore, RM, when used directly in riparian zones at a suitable concentration, is a novel and low cost additive material that can be used to remove phosphorus from reclaimed water.