The feasibility of using floating treatment wetlands (FTWs) to treat runoff typical of commercial nurseries was investigated using two 8-week trials with replicated mesocosms. Plants were supported by Beemat rafts. Five monoculture treatments of Agrostis alba (red top), Canna × generalis 'Firebird' (canna lily), Carex stricta (tussock sedge), Iris ensata 'Rising Sun' (Japanese water iris), Panicum virgatum (switchgrass), two mixed species treatments, and an unplanted control were assessed. These plant species are used for ornamental, wetland, and biofuel purposes. Nitrogen (N) and phosphorus (P) removals were evaluated after a 7-day hydraulic retention time (HRT). N removal (sum of ammonium-N, nitrate-N, and nitrite-N) from FTW treatments ranged from 0.255 to 0.738 g•m −2 •d −1 (38.9 to 82.4% removal) and 0.147 to 0.656 g•m −2 •d −1 (12.9 to 59.6% removal) for trials 1 and 2, respectively. P removal (phosphate-P) ranged from 0.052 to 0.128 g•m −2 •d −1 (26.1 to 64.7% removal) for trial 1, and 0.074 to 0.194 g•m −2 •d −1 (26.8 to 63.2% removal) for trial 2. Panicum virgatum removed more N and P than any other FTW treatment and the control in both trials. Results show that species selection and timing of FTW harvest impact the rate and mass of nutrient remediation. FTWs can effectively remove N and P from runoff from commercial nurseries.