If you went swimming or boating in the Hudson River Estuary (HRE) today, with current water quality practices, I could only tell you tomorrow if the water was safe for that interaction. This is not a very satisfying or useful solution for the public, environmental managers, and other stakeholders. One way to tackle this issue of delayed sewage contamination warning is by creating predictive models that are grounded in experimental data. Effectively predicting water quality and human health risks associated with sewage-derived microbes requires understanding the fate and transport of these contaminants. In my research, I examine how particle attachment of sewage bacteria alters their dynamics upon discharge to surface waters. In this presentation, I will share some recent modeling and experimental work on how particle association alters persistence of the fecal indicator bacteria Enterococcus sp. and of 2 genera with fecal pathogens: Salmonella sp. and Shigella sp. While the HRE is safe to swim (bacterially-speaking) roughly 80% of the time, it is important to develop methods for minimizing human health risk in the remaining time when water quality can be very poor.