Large-scale crustal thickening by tectonic and/or magmatic processes can lead to various complex patterns of multi-layered continental crust. It is well-known from one-dimensional thermal modeling that variations in total crustal thickness, mantle lithospheric thickness, thermal conductivities of the crust, and bulk radiogenic heat production of the crust will lead to variable geotherms in such heterogeneously stacked crust. By systematically changing the above parameters, we illustrate that variations on the order of 100–500°C will result at a depth of 30 km. Specifically, we show that geotherms are also strongly dependent on the pattern of vertical interlayering. Assuming a crustal structure composed of idealized granodioritic/gabbroic or granodioritic/dioritic compositional layer sequences, it can be shown that such gravitationally unstable, stacked, multi-layered continental crust can lead to temperature variations in geotherms of comparable magnitude as for the above parameters. Geotherms exhibiting the highest temperatures at a given depth are characteristic for gravitationally unstable structures in which the bulk of the granodioritic rocks underlie dioritic or gabbroic rocks. Thus a strong positive correlation between temperature and the estimated degree of gravitational instability of the multilayered crust is indicated. It is argued that the lowering of the viscosity of rocks with increasing temperature after tectonic or magmatic stacking will set the stage for processes of gravitational redistribution and buoyant exhumation of high-grade metamorphic rocks. Prograde changes in metamorphic mineral assemblages and partial melting during thermal relaxation after stacking provide positive feed-back mechanisms to enhance the possibility of gravitational redistribution. In keeping with the published results of Babeyko & Sobolev (2001) and Arnold et al. (2001), we find that gravitational overturn can be triggered only when external tectonic forces are active after stacking. Time-scales of 10 to 100 Myr are indicated for differential movement of rock masses on a kilometer-scale when the viscosity of crustal rocks is lowered to n × 1021 Pa·s, but may be considerably less if zones of tectonic weakness in the crust lead to a further local decrease in effective rock viscosity.