A replicated column trial was conducted to evaluate the potential for the phosphorus saturation index (PSI) to predict P movement in bioretention soil mixtures (BSMs). The impact of compost feedstock on BSM performance was also evaluated. Three composts (biosolids/yard, yard/food waste, and manure/sawdust) were each brought to PSI values of 0.1, 0.5, and 1.0 through the addition of Fe-based water treatment residuals (WTRs) to lower the PSI and P salts to increase the PSI. A synthetic stormwater solution was used for 12 leaching events. The PSI predicted total and dissolved P concentrations in column leachate. All composts removed P at PSI 0.1. All composts were a source of P for the higher PSI values tested, with P concentrations in the leachate decreasing over time. Ammonia and nitrate from all treatments decreased over time, with all treatments showing effective N removal. Copper removal (total and dissolved) was >90% for all treatments, with the highest removal observed at PSI 0.1 for all composts. Zinc removal (total) was also greatest in the 0.1 PSI for all composts. At PSI 0.5 and 1.0, the biosolids/yard compost was less effective than the other materials at removing Zn, with a removal efficiency of approximately 50%. Infiltration rates were similar across all treatments and ranged from 0.44 ± 0.1 cm min in the manure/sawdust at PSI 0.1 to 3.8 ± 2.8 cm min in the food/yard at PSI 1.0. Plant growth in the manure/sawdust compost was reduced in comparison to the other composts tested across all PSI levels. The results of this study indicate that the PSI may be an effective tool for predicting P movement in bioretention systems. Compost feedstock does not indicate the ability of composts to filter contaminants filtration, with all composts tested showing high contaminant removal.