Few studies have characterized reactive media for phosphorus (P) removal in passive treatment systems in terms of both batch and continuous flow experiments. This study uses basic oxygen steel slag (BOS) from a U.K. feedstock. Batch experiments demonstrated the effective removal of phosphorus with varying initial pH, initial P concentration, clast size, and ionic strength to represent environmental conditions. Continuous flow column experiments, operated for 406 days, with an influent P concentration of 1-50 mg/L (typical of domestic and dairy parlour waste) achieved removal of up to 62%; a second set of column experiments running for 306 days with an influent P concentration of 100-300 mg/L achieved a maximum effective removal of 8.39 mg/g. This figure is higher than that for other slags reviewed in this study (e.g., EAF Slag 3.93 mg/g and NZ melter slag 1.23 mg/g). XRD, E-SEM, and EDX data provide evidence for a sequential series of increasingly less soluble P mineral phases forming on the BOS surface (octa-calcium phosphate, brushite, and hydroxylapatite),which suggests that BOS may be a suitable substrate in passive treatment systems, providing a long-term P removal mechanism.