Abstract The West Siberian Basin (WSB) records a detailed history of Permo-Triassic rifting, extension and volcanism, followed by Mesozoic and Cenozoic sedimentation in a thermally subsiding basin. Sedimentary deposits of Permian age are absent from much of the basin, suggesting that large areas of the nascent basin were elevated and exposed at that time. Industrial seismic and well log data from the basin have enabled extension and subsidence modelling of parts of the basin. Crustal extension (β) factors are calculated to be in excess of 1.6 in the northern part of the basin across the deep Urengoy graben. 1-D backstripping of the Triassic to Cenozoic sedimentary sequences in this region indicates a period of delayed subsidence during the early Mesozoic. The combination of elevation, rifting and volcanism is consistent with sublithospheric support, such as a hot mantle plume. This interpretation accords with the geochemical data for basalts from the Siberian Traps and the West Siberian Basin, which are considered to be part of the same large igneous province. Whilst early suites from Noril'sk indicate moderate pressures of melting (mostly within the garnet stability field), later suites (and those from the West Siberian Basin) indicate shallow average depths of melting. The main region of magma production was therefore beneath the relatively thin (ca. 50–100 km) lithosphere of the basin, and not the craton on which the present-day exposure of the Traps occurs. The indicated uplift, widespread occurrence of basalts, and short duration of the volcanic province as a whole are entirely consistent with published models involving a mantle plume. The main argument against the plume model, namely lack of any associated uplift, appears to be untenable.