Published Holocene relative sea-level (RSL) curves for the U.S. Gulf Coast are in mutual conflict, with some characterized by a smooth RSL rise akin to widely accepted eustatic sea-level curves versus others, including several recent ones, that are characterized by a conspicuous "stair-step" pattern with prolonged (millennium-scale) RSL stillstands alternating with rapid (meter-scale) rises. In addition, recent work in Texas and Alabama has revitalized the notion of a middle Holocene RSL highstand, estimated at 2 m above present mean sea level. An extensive sampling program in the Mississippi Delta (Louisiana) focused on the collection of basal peats that accumulated during the initial transgression of the pre-existing, consolidated Pleistocene basement. We used stable carbon isotope ratios to demonstrate that many of these samples accumulated in environments affected by frequent saltwater intrusion in the <30 cm zone between mean spring high water and mean sea level, and we selected plant macrofossils that were subjected to AMS 14C dating. Nearly 30 sea-level index points from a ∼20 km2 study area on the eastern margin of the delta suggest that RSL rise followed a relatively smooth trend for the time interval 8000-3000 cal yr B.P., thus questioning the occurrence of major RSL stillstands alternating with abrupt rises. Given the narrow error envelope defined by our data set, any sea-level fluctuations, if present, would have amplitudes of <1 m. Although a true middle Holocene highstand never occurred in the Mississippi Delta, the high level of detail of our time series enables a rigorous test of this hypothesis. Correction of our data set for a hypothetical tectonic subsidence rate of 1.1 mm yr-1 (assuming a constant subsidence rate compared to the tectonically relatively stable adjacent coast of Texas) leads to sea levels near 2 m above present during the time interval 6000-4000 cal yr B.P. However, this model also implies a RSL position near -2 m around 8000 cal yr B.P., which is inconsistent both with data of this age from Texas, as well as with widely accepted sea-level data from elsewhere. We therefore conclude that a middle Holocene highstand for the U.S. Gulf Coast is highly unlikely, and that the entire area is still responding glacio-isostatically, by means of forebulge collapse, to the melting of the Laurentide Ice Sheet.