Two paradoxical geological features of the Himalaya are the syn-convergence extension and the inverted metamorphic isograds observed in the crystalline core zone of this orogen. This High Himalayan Crystalline Sequence corresponds to an up to 40 km thick sequence of amphibolite to granulite facies gneiss, bounded by the Main Central Thrust at the base, and by the extensional faults of the South Tibetan Detachment System at the top. Geochronological and structural data demonstrate that coeval movements along both the Main Central Thrust and South Tibetan Detachment System during Early to Middle Miocene times were related to a tectonically controlled exhumation of these high-grade metamorphic rocks. The High Himalayan Crystalline Sequence systematically shows an inverted metamorphic zonation, generally characterized by a gradual superposition of garnet, staurolite, kyanite, sillimanite + muscovite and sillimanite + K-feldspar isograds, from the base to the top of the unit. Recent kinematic flow analyses of these metamorphic rocks demonstrate the coexistence of both simple shear and pure shear during the ductile deformation. The simple shear component of such a general non-coaxial flow could explain a rotation of isograds, eventually resulting in an inversion. The pure shear component of the flow implies a thinning of the metamorphic sequence that must be balanced by a perpendicular stretching of the unit parallel to its boundaries. Inasmuch as seismic data show that both the Main Central Thrust and South Tibetan Detachment System converge at depth, a thinning of the wedge-shaped High Himalayan Crystalline Sequence should induce a ductile extrusion of these high-grade rocks toward the surface. Rapid extension at the top of the sequence could thus be the consequence of a general shear extrusion of this unit relative to its hanging wall. Moreover, this extensional movement should decrease with depth to become zero where the boundaries of the unit meet, accounting for the paradoxical convergence of the South Tibetan Detachment System toward the Main Central Thrust. Furthermore, a general flow combining simple shear and pure shear can reconcile inverted isograds with the lack of inverted pressure field gradient across the High Himalayan Crystalline Sequence, despite an intense non-coaxial deformation. In good agreement with the seismic, kinematic and P–T–t constraints on the Himalayan tectono-thermal evolution, general shear extrusion provides a consistent model accounting for both inverted isograds and rapid extension in a compressional orogenic setting.