Appalachian Carboniferous stratigraphy remains controversial partly because of unclear petrogenetic, depositional, and stratigraphic relationships displayed by Pottsville sandstones, which have been interpreted as being of either alluvial or barrier-beach origin. Two associated sandstones of the New River Formation of southern West Virginia provide the setting for a detailed examination of this problem. The Lower Raleigh lithic arenite consists of elongate and lobate, dip-oriented bodies which display erosional bases, interfinger laterally with shales, fine upward at the top, and are unidirectionally trough cross-bedded basinward. The Upper Raleigh quartz arenite consists of elongate, dip-oriented bodies updip in the basin and elongate and lobate, strike-oriented bodies downdip. The quartz arenites display erosional bases, interfinger laterally with shales, fine upward at the top, and are unidirectionally planar cross-bedded basinward. Evidence suggests that Lower Raleigh detritus was derived from a southeasterly metamorphic-sedimentary source terrane and deposited in high-constructive lobate deltas. Upper Raleigh quartzose detritus may have been derived from coarse grained alluvial facies equivalents of Lower Raleigh deltaic facies and deposited in deltas whose margins were modified by marine processes. It appears that the Lower and Upper Raleigh sandstones are not deltaic and barrier-tidal facies equivalents, thus negating gross regional stratigraphic relationships implied by the barrier model. Additionally, these conclusions suggest that use of quartz arenite as a unique lithogenetic rock type for predictive purposes is prone to serious error.