Biorention cells or rain gardens are used to filter runoff before it reaches the storm sewer. Concern has been raised that compost in the filter soil may generate nutrient load rather than reducing nutrient load of runoff passed through the biorentention cell. The purpose of this study was to determine how nutrient filtering would be affected by the following: sieved versus unsieved soil in the mix, adding perlite and biochar to sorb P, layered vs. mixed material, and different mixture proportions. Mixtures were packed into PVC columns, and disk infiltrometers were used for the initial wetting. Then prairie mixtures (Buffalo grass, Buchloe dactyloides (Nutt.) Engelm., blue grama grass, Bouteloua gracilis H.B.K. and purple coneflower, Echinacea purpurea L.) were planted to the columns and grow lights established. Columns were watered with a solution that contained suspended solids as well as fertilizer. For the initial watering the columns without soil (sand and compost, with or without perlite and biochar), had significantly faster ponded infiltration (627 and 490 mm/h) than the columns with sand, compost, and soil (whether sieved, 157 mm/h, or not, 139 mm/h). The columns with only sand and compost had significantly faster ponded infiltration than than the columns layered with compost and sand mixed with perlite and biochar (415 mm/h). Effluent during initial wetting, growing, and final infiltration will be analyzed for suspended solids, total P, and nitrate-N. Results of this study will be used to further refine the recommendations used statewide for the filter soil media.