Salt-marsh sediments provide precise and near-continuous reconstructions of Common Era relative sea level (RSL). However, organic and low-density salt-marsh sediments are prone to compaction processes that cause post-depositional distortion of the stratigraphic column used to reconstruct RSL. We compared two RSL reconstructions from East River Marsh (Connecticut, USA) to assess the contribution of mechanical compression and biodegradation to compaction of salt-marsh sediments and their subsequent influence on RSL reconstructions. The first, existing reconstruction (‘trench’) was produced from a continuous sequence of basal salt-marsh sediment and is unaffected by compaction. The second, new reconstruction is from a compaction-susceptible core taken at the same location. We highlight that sediment compaction is the only feasible mechanism for explaining the observed differences in RSL reconstructed from the trench and core. Both reconstructions display long-term RSL rise of ∼1 mm/yr, followed by a ∼19th Century acceleration to ∼3 mm/yr. A statistically-significant difference between the records at ∼1100 to 1800 CE could not be explained by a compression-only geotechnical model. We suggest that the warmer and drier conditions of the Medieval Climate Anomaly (MCA) resulted in an increase in sediment compressibility during this time period. We adapted the geotechnical model by reducing the compressive strength of MCA sediments to simulate this softening of sediments. ‘Decompaction’ of the core reconstruction with this modified model accounted for the difference between the two RSL reconstructions. Our results demonstrate that compression-only geotechnical models may be inadequate for estimating compaction and post-depositional lowering of susceptible organic salt-marsh sediments in some settings. This has important implications for our understanding of the drivers of sea-level change. Further, our results suggest that future climate changes may make salt marshes more susceptible to the impacts of RSL rise by enhancing sediment compressibility. We stress, however, that the cause of the softening remains enigmatic. Until this is better constrained, it is premature to widely extrapolate our findings to existing core-based reconstructions of Holocene RSL.