As a contribution to the plume-nonplume debate we review the tectonic setting in which huge volumes of monotonous tholeiite of the Columbia River flood basalt province of the Pacific Northwest, USA, were erupted. We record the timescale and the locations of these eruptions and estimates of individual eruption volumes, and we discuss the mechanisms of sheet-flow emplacement, all of which bear on the ultimate origin of the province. An exceptionally large chemical and isotopic database is used to identify the various mantle sources of the basalt and their subsequent evolution in large lower-crustal magma chambers. We conclude by discussing the available data in light of the various deep-mantle plume and shallow-mantle models recently advocated for the origin of this flood basalt province, and we argue that the mantle plume model best explains the eruption of such an exceptionally large volume of tholeiitic basalt within such an unusually short period and within such a restricted area.