Submerged aquatic vegetation (SAV) provides many important ecosystem functions, but SAV has been significantly reduced in many estuaries. We used spatial–statistical models to identify estuarine shoreline characteristics that explain variations in SAV abundance among subestuaries of the Chesapeake Bay and mid-Atlantic Coastal Bays. We summarized digital spatial data on shoreline construction, shoreline land use, physical characteristics, watershed land cover, and salinity for each subestuary. We related SAV abundance to shoreline characteristics and other stressors using univariate regression and multivariate models. The strongest univariate predictors of SAV abundance were percent shoreline forest, percent shoreline marsh, the percentage of shoreline that is 5–10 m tall, percent riprap, the percentage of subestuary area <2 m deep, percent herbaceous wetland, and percent shrubland. Shoreline marsh, bulkhead, and shoreline forest had different effects on SAV in different salinity zones. Percent riprap shoreline was the most important variable in a regression tree analysis of all the subestuaries, and percent deciduous forest in the watershed was the most important variable in a separate regression tree analysis on the mesohaline subestuaries. Subestuaries with <5.4 % riprap followed a significantly different temporal trajectory than those with >5.4 % riprap. SAV abundance has increased steadily since 1984 in subestuaries with <5.4 % riprap, but has not increased since 1996–1997 in subestuaries with >5.4 % riprap. Some shoreline characteristics interact with larger-scale factors like land cover and salinity zone to affect the distribution of SAV, while the effects of other shoreline characteristics are consistent among subestuaries with different salinities or local watershed land covers. Many shoreline characteristics can be controlled by management decisions, and our results help identify factors that managers should consider in efforts to increase SAV abundance.