Abstract The Pakistan part of the Himalaya has major differences in tectonic evolution compared with the main Himalayan range to the east of the Nanga Parbat syntaxis. There is no equivalent of the Tethyan Himalaya sedimentary sequence south of the Indus–Tsangpo suture zone, no equivalent of the Main Central Thrust, and no Miocene metamorphism and leucogranite emplacement. The Kohistan Arc was thrust southward onto the leading edge of continental India. All rocks exposed to the south of the arc in the footwall of the Main Mantle Thrust preserve metamorphic histories. However, these do not all record Cenozoic metamorphism. Basement rocks record Paleo-Proterozoic metamorphism with no Cenozoic heating; Neo-Proterozoic through Cambrian sediments record Ordovician ages for peak kyanite and sillimanite grade metamorphism, although Ar–Ar data indicate a Cenozoic thermal imprint which did not reset the peak metamorphic assemblages. The only rocks that clearly record Cenozoic metamorphism are Upper Paleozoic through Mesozoic cover sediments. Thermobarometric data suggest burial of these rocks along a clockwise pressure–temperature path to pressure–temperature conditions of c. 10–11 kbar and c. 700°C. Resolving this enigma is challenging but implies downward heating into the Indian plate, coupled with later development of unconformity parallel shear zones that detach Upper Paleozoic–Cenozoic cover rocks from Neoproterozoic to Paleozoic basement rocks and also detach those rocks from the Paleoproterozoic basement.