Bioretention systems, also known as rain gardens or bioinfiltration, offer a promising approach to nature-based water infrastructure for managing, treating, and potentially reusing stormwater. Gaining insight into their sustainability implications would enhance the design and operation of these systems for non-potable stormwater reuse. However, questions remain regarding associated potential human health risks and environmental impacts. This study evaluates environmental and human health risks from urban stormwater reuse using bioretention mechanistic modeling paired with quantitative microbial risk assessment (QMRA) and life cycle assessment (LCA) under uncertainty. Design alternatives are compared for treating stormwater polluted with heavy metals, nutrients, and pathogens. The study investigates the influence of different stochastic configurations of bioretention media (including sand, compost, woodchips, and loam) on removal of pathogens, heavy metals (zinc and copper), and nutrients (nitrogen and phosphorus) to ultimately model human health and environmental impacts. Modeled health risks from two pathogens (Campylobacter jejuni, Cryptosporidium) after consumption of irrigated salad, washed salad, accidental water ingestion during urban farming, and recreational swimming were analyzed. In comparison to influent concentrations, media composition had minimal influence on the modeled risk. Under uncertainty (using Monte Carlo), microbial growth in the washed salad influenced the high median annual risks from reuse (4.47 x 10-4 from Campylobacter jejuni and 1.67 x 10-3 from Cryptosporidium) compared to recreational swimming which had the lowest median annual risk. These findings will be combined with results from the life cycle assessment (with uncertainty and sensitivity analyses) to explore sustainability trade-offs and comprehensively guide decision-making for bioretention design.