Abstract Deep seismic data that have been shot across the world’s passive margins make us reflect that much of the subsidence that post-dates major rifting and continental separation is not thermal in origin, but structural, associated with the localization of extensional displacement on a major fault or shear zone along the subcontinental Moho. Displacement surfaces of this kind have been called ‘exhumation faults’ ( Manatschal et al ., 2007 ),‘detachment faults’ ( Manatschal and Lavier, 2010 ; Reston and McDermott, 2011 ), and ‘outer marginal detachments’ (Pindell et al ., in prep., and this meeting). On non-volcanic margins they may exhume the Moho at the sea bed; on volcanic margins they may represent magma welds (Pindell, this meeting). We believe that the subsidence is structural collapse of the upper part of the continental crust. On volcanic margins it is probably associated with the pinching out (boudinage) of the Lower Crust so that the Upper crust effectively collapses onto the mantle. On volcanic margins with SDRs, the collapse of both the continental edge and the lava flows (SDRs) that overlie it may be due to accommodation space being created along an evacuating magma weld. We believe that this sort of collapse is rapid, far quicker than thermal subsidence, and attempt to support the idea by examples from the Gulf of Mexico, Brazil, the Alps, and the Red Sea. The recognition of rapid collapse is not new. It is well described in classic stratigraphic literature in the Alps and elsewhere. Here we argue that its occurrence is extremely widespread, but is commonly overlooked.