In New South Wales (NSW) Australia, the recent introduction of legislation to control runoff and charge for water used in agricultural production has encouraged commercial plant nurseries to collect and recycle their irrigation drainage. Runoff from a nursery typically contains around 6 mg/L TN (> 70% as NO3), 0.5 mg/L TP (> 50% as P04), and virtually no organic matter (BOD < 5 mg/L; DOC < 20 mg/L). As a result, algal blooms frequently occur in storage dams. This paper describes a study evaluating the effectiveness of subsurface flow wetlands in the removal of nutrients from nursery runoff on the sub-tropical northern coast of NSW, Australia. Four experimental subsurface flow wetlands (1 m x 4 m x 0.5 m water depth) were planted with Phragmites australis in April 1999. TN and TP load removals were > 84% and > 65% respectively at HRTs of between 5 and 2 days, with the majority of out-flowing TN and TP being organic in form. Internal generation of organic N and P resulted in persistent background levels of 0.45 mg/L TN and 0.15 mg/L TP in the reed bed effluent. TN, NH4 and TP removal was affected by HRT (P < 0.05). Greater than 90% load removal of NH4, NO2, NO3 and Ortho-P was achieved at all HRTs, with outlet concentrations generally < 0.01 mg/L for all. For TN, a strong relationship existed between removal rate (g/m2/day) and loading rate (r2 = 0.995), while a weaker relationship existed for TP (r2 = 0.47). It is estimated that a 1 ha nursery would require a reed bed area of 200 m2 for a 2 day HRT.