Stormwater best management practices (BMPs) are implemented to reduce microbial pollution in runoff, but their removal efficiencies differ. Enhanced BMPs, such as those with media amendments, can increase removal of fecal indicator bacteria (FIB) in runoff from 0.25-log10 to above 3-log10; however, their implications for watershed-scale management are poorly understood. In this work, a computational model was developed to simulate watershed-scale bacteria loading and BMP performance using the Ballona Creek Watershed (Los Angeles County, CA) as a case study. Over 1400 scenarios with varying BMP performance, percent watershed area treated, BMP treatment volume, and infiltrative capabilities were simulated. Incremental improvement of BMP performance by 0.25-log10, while keeping other scenario variables constant, reduces annual bacterial load at the outlet by a range of 0-29%. In addition, various simulated scenarios provide the same FIB load reduction; for example, 75% load reduction is achieved by diverting runoff from either 95% of the watershed area to 25 000 infiltrating BMPs with 0.5-log10 removal or 75% of the watershed area to 75 000 infiltrating BMPs with 1.5-log10 removal. Lastly, simulated infiltrating BMPs provide greater FIB reduction than noninfiltrating BMPs at the watershed scale. Results provide new insight on the trade-offs between BMP treatment volume, performance, and distribution.