Abstract The Platte River of Nebraska, central USA, has been used by sedimentologists as a type example of a sandy, braided river system for some forty years. Despite this designation, however, the details of the river’s subsurface stratigraphy have never been evaluated, and so the extant facies model for the Platte River remains unvalidated. This paper analyzes the Holocene stratigraphy of the central and lower Platte River. It utilizes newly acquired data from wireline logs, sediment cores, aerial photography, and ground-penetrating radar (GPR) traverses to characterize subsurface geology, in addition to surface sediment and geomorphic mapping, sampling, and analysis of serial aerial imagery. Cores and GPR profiles show that the lithosome of the active Platte is 6–8 m in thickness and consists predominantly of medium sand. Although macroform features are common on the surface of the modern (active) central and lower Platte River in Nebraska, the shallow subsurface stratigraphy is dominated by mesoform cross-bedding and planar stratification overlying channel scour fills. This differs markedly from the vertical succession predicted by the Platte River facies model of Miall (1977). Data from inactive portions of the central and lower Platte River, however, show macroform cross-bedding indicative of the downstream and cross-stream migration of linguoid and transverse bars, which would be predicted by the published model. The typical architecture of the Holocene channel-belt sand body in the areas studied thus comprises a lower unit dominated by macroform-scale cross-bedding, overlain by an upper unit dominated by mesoform-scale cross-bedding and scour fills. Furthermore, time-series analysis of aerial photography of the river in the vicinity of Grand Island, Nebraska, shows that: (1) the river has become dominated by aggressively vegetated subaerially exposed channel-bed domains (termed “vegetated channel islands”) that superficially resemble bars; and (2) that this stretch is no longer dominated by downstream-accreting macroform bars, which were abundant in the 1960s and 1970s. Contraction of the active channel in this region has accompanied these changes, somewhat independently of discharge variation. New vertical sequence facies models are presented to account for these differences. Our data show the futility of seeking to characterize a fluvial system by means of a single vertical sequence model, and furthermore indicate that the Holocene facies architecture of the sandy, braided Platte River varies both in space and in time. The dominance of mesoform-scale structure in the shallow subsurface of the central and lower Platte River likely reflects the ongoing process of gradual abandonment of the system. It further highlights the need for caution in interpreting subsurface structure from surface geomorphology in fluvial systems.