Sediment resuspension and related increases in turbidity in shallow coastal bays are strongly controlled by local bed properties. However, knowledge of bed properties in coastal bays is typically sparse at best. In this study, we developed a method to estimate the spatial distribution of bed properties in shallow coastal bays using a combination of bed sediment measurements and residence time calculations that requires neither extensive dedicated modeling nor extensive sampling. We found a strong relationship between water residence times derived from a coastal hydrodynamic model and observed bed grain size fractions in a system of coastal bays and used that relationship to transform maps of residence time to maps of grain size fractions throughout the bays. Because grain‐size fractions are related to other bed properties such as organic fraction, permeability and cohesion, these maps provide valuable information for habitat studies as well as morphodynamic modeling. We used our maps of grain size distributions to initialize a 2‐month‐long model simulation of currents, waves and suspended sediment forced with measured wind and tides. Spatial variations in suspended sediment concentration (SSC) reflected spatial gradients in sand and mud abundance in the bed. Lower SSC in sandier regions of the bays, near barrier islands and inlets, resulted in higher benthic light availability but lower sediment supply for deposition on back‐barrier marshes. Higher SSC in more landward, muddier regions resulted in greater light attenuation and sediment availability for deposition on mainland fringing marshes. The proposed methodology facilitates quantification of these bed‐dependent spatial variations in SSC.