Some mechanisms in favor of later regional and earlier graben type subsidence at passive continental margins are reviewed. Gravity loading can explain thick sediment piles beneath some deltas but fails to account for most shelf subsidence except as a contributory factor. Heating of the lithosphere resulting in crustal thinning by uplift and erosion and/or increased lower crustal density by metamorphism or intrusion can account for shelf subsidence of up to about 3 to 5 km after sediment loading. Seaward creep of continental crustal material may cause substantial subsidence owing to thinning of the crust beneath the shelf. Thinning of the crust by "necking" prior to continental splitting appears only viable in a narrow graben setting where igneous intrusion can extend the pper crust. Graben subsidence prior to splitting can best be explained by the wedge subsidence mechanism applied to the upper 10 km of the crust rather than to the whole crust. Most of these mechanisms probably contribute to observed subsidence although their relative importance is not yet clear.