Abstract. Although tide gauges are the primary source of data used to calculate multi-decadal- to century-scale rates of relative sea-level change, we question the usefulness of tide-gauge data in rapidly subsiding low-elevation coastal zones (LECZs). Tide gauges measure relative sea-level rise (RSLR) with respect to the base of associated benchmarks. Focusing on coastal Louisiana, the largest LECZ in the United States, we find that these benchmarks (n=35) are anchored an average of 21.5 m below the land surface. Because at least 60 % of subsidence occurs in the top 5 m of the sediment column in this area, tide gauges in coastal Louisiana do not capture the primary contributor to RSLR. Similarly, global navigation satellite system (GNSS) stations (n=10) are anchored an average of > 14.3 m below the land surface and therefore also do not capture shallow subsidence. As a result, tide gauges and GNSS stations in coastal Louisiana, and likely in LECZs worldwide, systematically underestimate rates of RSLR as experienced at the land surface. We present an alternative approach that explicitly measures RSLR in LECZs with respect to the land surface and eliminates the need for tide-gauge data in this context. Shallow subsidence is measured by rod surface-elevation table–marker horizons (RSET-MHs) and added to measurements of deep subsidence from GNSS data, plus sea-level rise from satellite altimetry. We show that for an LECZ the size of coastal Louisiana (25 000–30 000 km2), about 40 RSET-MH instruments suffice to collect useful data. Rates of RSLR obtained from this approach are substantially higher than rates as inferred from tide-gauge data. We therefore conclude that LECZs may be at higher risk of flooding within a shorter time horizon than previously assumed.