Accelerated sea-level rise (SLR) poses the threat of salt marsh submergence, especially in marshes that are relatively low-lying. At the same time, restoration efforts are producing new low-lying marshes, many of which are thriving and avoiding submergence. To understand the causes of these different fates, we studied two Long Island Sound marshes: one that is experiencing submergence and mudflat expansion, and one that is undergoing successful restoration. We examined sedimentation using a variety of methods, each of which captures different time periods and different aspects of marsh elevation change: surface-elevation tables, marker horizons, sediment cores, and sediment traps. We also studied marsh hydrology, productivity, respiration, nutrient content, and suspended sediment. We found that, despite the expansion of mudflat in the submerging marsh, the areas that remain vegetated have been gaining elevation at roughly the rate of SLR over the last 10 years. However, this elevation gain was only possible thanks to an increase in belowground volume, which may be a temporary response to waterlogging. In addition, accretion rates in the first half of the twentieth century were much lower than current rates, so century-scale accretion in the submerging marsh was lower than SLR. In contrast, at the restored marsh, accretion rates are now averaging about 10 mm yr⁻¹ (several times the rate of SLR), much higher than before restoration. The main cause of the different trajectories at the two marshes appeared to be the availability of suspended sediment, which was much higher in the restored marsh. We considered and rejected alternative hypotheses, including differences in tidal flooding, plant productivity, and nutrient loading. In the submerging marsh, suspended and deposited sediment had relatively high organic content, which may be a useful indicator of sediment starvation.