Efforts have been made in U.S. wetlands to eradicate the invasive grass Phragmites australis. But eradication of Phragmites does not always lead to the return of native plants. This research investigated native vegetation recolonization across 12 tidal wetland study sites in the Chesapeake Bay watershed and tested the potential of planting perennial native wetland species to accelerate site recovery following Phragmites removal. Our study found that site salinity was a dominant driver of plant recolonization rate: low salinity sites (0.5-3 ppt) had, on average, 4.3x greater aboveground biomass and 2.5x higher vegetation cover than brackish sites (5-9 ppt) two or more growing seasons after Phragmites removal. The composition of returning plant species also differed by salinity, with a higher abundance of annuals and fewer graminoids at low-salinity sites. Site hydrology also influenced native plant recolonization—more frequently flooded sites had lower aboveground biomass of native vegetation two or more years following Phragmites removal. Experimental planting had variable results, with high die-off at several sites, but showed potential to accelerate vegetation recovery at brackish sites in the first years after Phragmites removal—plots with transplants at brackish sites had 17.5x, 2.4x, and 1.5x higher plant cover than unplanted plots in years one, two, and three, respectively, after planting. All sites had some amount of native vegetation recovery within three to four years following Phragmites removal, suggesting that native planting may not be necessary for many tidal wetland sites. Sites with especially high salinities and flooding frequencies may benefit the most from plantings, as larger plants may be able to survive in conditions that are not favorable for seedling emergence. In a mesocosm experiment, we planted six different clonal wetland species in a sand-vermiculite mixture at three different elevations in a tidal creek on the Rhode River in Maryland, USA. We found that peak plant biomass in the sandy substrate occurred at lower elevations and higher flooding frequencies than is typical in marsh environments and than was observed in other mesocosm experiments with organic soils. In well-drained, sandy substrates, wetland plants may benefit from more frequent tidal pulses, likely due to increased water supply and nutrient flux. This has implications for wetland-restoration practitioners using dredged sand to create or elevate tidal wetlands, as wetland species may grow at different elevations and flooding frequencies in these conditions than in a typical tidal marsh with organic soils.