It is widely assumed that the India-Asia collision initiated at 55 5 Ma along the Indus-Yarlung suture zone. Paradoxically, however, the majority of tectonic responses to collision as well as the accelerated faunal exchange between the two continents did not occur until circa 35-30 Ma and intense deformation and uplift occurred along the southern side of the Tibetan Himalaya rather than the Indus-Yarlung suture zone. Additionally, this widely accepted scenario requires large-scale subduction of buoyant continental crust into the dense mantle, which is inconsistent with tectonic reconstructions of Gondwana as well as the general knowledge of physics. Here we present evidence from crustal volume budget calculations to show that (1) at 55 5 Ma 2,100 840 km of oceanic lithosphere existed between India and the Lhasa Block and hence India-Asia collision is unlikely to have occurred at that time, (2) substantial India-Asia collision should have occurred after this large region of oceanic lithosphere had been consumed by subduction, along the south side of Tibetan Himalaya diachronously at 32 2 Ma in the west, 23 2 Ma in the east and 18 +/- 2 Ma in the center, and (3) no large-scale subduction of continental crust occurred during India-Asia collision. These findings resolve the problematic issues that have not been satisfactorily explained by many existing models for the system and suggest that the evolution of the Himalaya-Tibet orogen and many existing geodynamic models of India-Asia collision requires critical reassessment. Plain Language Summary India-Asia collision, which built the spectacular Himalaya-Tibet orogen, is widely considered to have commenced at circa 55 Ma along the Indus-Yarlung suture zone, but apparent paradoxes have made this widely accepted scenario difficult to understand. Based upon an analysis from crustal volume budgets, a common method for the restoration of tectonic deformation, we suggest that the India-Asia collision is unlikely to have occurred at circa 55 Ma along the Indus-Yarlung suture zone but instead most likely commenced after circa 32 Ma along the south side of Tibetan Himalaya. Our findings provide reasonable explanations for the problematic issues that cannot be explained by the many existing models that propose that India-Asia collision occurred at circa 55 Ma.