Escherichia coli release from streambed sediments may substantially affect microbial water quality. Models of E. coli release and transport commonly use a single set of parameters for the whole stream or reservoir, yet little is known about the magnitude and sources of the variability of parameters of the streambed bacteria release. The objectives of this work were: (a) to obtain and compare parameters of streambed E. coli resuspension in three stream reaches with distinctly different bottom sediment textures, and (b) to see whether the modeling of streambed E. coli resuspension with reach-specific parameters could provide substantially better accuracy than modeling with a single set of parameters. Sediment particle size distributions and the streambed E. coli concentrations were measured along a first-order creek in the USDA-ARS OPE3 experimental watershed in Maryland. Afterwards, 80m³ of water were released into the creek at a rate of 60L per second in four equal allotments separated by 1–3min intervals. Flow rates and E. coli concentrations were monitored with automated samplers at the ends of the three reaches with a total length of 630m. A high concentration of streambed E. coli (“hotspot”) resuspended within the first reach caused a pulse of high E. coli concentrations that propagated along the creek without substantial attenuation; inputs of sediment-borne E. coli from the next two reaches were relatively small. The E. coli transport model included one-dimensional Saint–Venant and advective–dispersive equations. The calibrated roughness coefficient values were comparable for the three reaches, whereas the critical stress and the entrainment rate differed among reaches by a half order and an order of magnitude, respectively. Overall, better accuracy was observed when the model contained reach-specific parameters. Additional research is needed to understand which and how sediment properties affect parameters of streambed E. coli release into the water column.