Across the world, cities are spending billions of dollars to manage urban runoff through decentralized green infrastructure (GI). This research uses an agent-based model to explore some of the physical, social, and economic consequences of one such urban GI programs. Using the Bronx, NY, as a case study, two alternative approaches to GI application are compared. The first (Model 1) mimics NYC's current GI program by opportunistically selecting sites for GI within the city's priority combined sewer watersheds; the second (Model 2) features a more spatially flexible approach to GI siting, in which the city attempts to maximize opportunities for co-benefits within the geographic areas considered in Model 1. The effects of both approaches, measured in terms of stormwater captured and co-benefits (e.g., carbon sequestered) provided, are tracked over 20-year simulations. While both models suggest it will be difficult to meet the citywide stormwater capture goals (managing the first 2.5 cm of rainfall from 10% of impervious surfaces) in the Bronx solely through public investment in GI, Model 2 shows that by integrating GI with other city initiatives (e.g., sustainability goals and resilience planning), synergistic outcomes are possible. Specifically, Model 2 produces stormwater capture rates comparable to those obtained under Model 1, but these rates are accompanied by elevated co-benefits for Bronx communities. The results are discussed in the context of future GI policy development in NYC. GI sites utilize soil, vegetation, and other natural features to retain, detain, and infiltrate urban runoff, diverting it from engineered collection systems. These processes help treat stormwater at the source, reducing pollutant loads, replenishing soil moisture, recharging aquifers, and mitigating local flooding (Narayan et al.