The problem of the observed very rapid advection of heat into metamorphic thrust stacks is reviewed. Conductive models relying on the thermal relaxation of a thickened crust will not produce the observed Barrovian (medium temperature, medium pressure) assemblages within some short-lived orogens (e.g., western Ireland and Timor). Studies of the rate and timing of metamorphic mineral growth suggest that this is commonly faster than predicted by thermal relaxation. Barrovian assemblages are localised in some orogens (e.g., the Alps) but extensive in others (e.g., the Himalayas). Metamorphic mineral growth brackets deformation; consequently, slow growth is inconsistent with the rapid uplift of many orogens. Thus, no single mechanism can account for the development of Barrovian assemblages during collisional orogeny. The only mechanisms that can supply large amounts of heat for regional metamorphism quickly (<10 Myr) are: rapidly thinning the lithosphere without stretching it (e.g., by plume thermal erosion, slab drop-off, or delamination); by emplacing magma into the crust (modest deep mafic underplate and (or) very large amounts of mafic and silicic magma emplaced into the middle and upper crust); or obducting hot nappes of arc with a thin ophiolite forearc (“hot iron” mechanism). Frictional and viscous heating produces local rapid heating but not fast regional heating. Back-arc or any kind of lithospheric extension increases the geothermal gradient and heat flow but does not heat rocks up. We suggest that magmatic advection of heat-associated lithospheric thinning or “hot iron” overthrusting of an arc/ophiolite are the primary sources of heat in short-lived orogens.
