In order to evaluate a model of Holocene shelf sediment distribution requiring a nearshore modern sand facies and an offshore relict sand facies, we have undertaken a textural reconnaissance of the Virginia-North Carolina Coast between Capes Henry and Hatteras. Grab samples were subjected to grain-size analysis by means of a modified Woods Hole Rapid Sediment Analyser. Textural provinces were erected with the aid of factor vector analysis. These include medium-grained sands of the beach and surf zones; seaward fining, fine-grained sands of the shore face, and heterogenous sands of the sea floor. In this latter province, grain size is controlled by a ridge and swale topography, with coarser sand on the crests. Coast-wise grain-size trends on the beach and shore face can be explained by assuming that wave heights increase toward the south, and that the Pleistocene sediment source is exposed higher on the shore face in the north than it is in the south. The shore face is retrograding, except in the vicinity of Diamond Shoals. There is textural evidence for a former Albemarle River channel, which bisects the study area. A model for sediment fractionation on a retreating barrier coast with low sediment input is proposed, based on studies which indicate that on such coasts: (1) barrier superstructures retreat more or less continuously by upper shore-face erosion and storm washover; and (2) lower shore-face erosion results in an equal-volume aggradation of the adjacent sea floor, and forms the leading edge of the Holocene transgressive sand sheet. The nearshore “modern” sands and offshore “relict” sands are both present in the study area, but the terms are unnecessarily restrictive. Both are “relict” in the sense of being derived from a Pleistocene substrate, and both are “modern” in the sense of having undergone adjustment to a modern hydraulic regime. While modern and relict are useful general terms, it is convenient in this area to refer to a Holocene barrier sand prism, versus a Holocene transgressive sand sheet.