The switch from I- to S-type magmatism in the Trans-Himalayan Batholith is one of the criteria used to determine when India and Eurasia collided. We present new 238U/ 206Pb SHRIMP data from the Karakorum Batholith, NW India. Our results suggest that an I-type hornblende-biotite granodiorite was emplaced at 31.4 ± 0.4 Ma and crosscut by S-type granite dykes at 18.0 ± 0.4 Ma. We interpret these data to indicate that volatile-fluxed melting of a mantle wedge above an active subduction zone and/or decompression melting of sub-arc asthenosphere was possible until at least 31 Ma. This interpretation uses the same rationale as previous workers have adopted to propose that the India-Eurasia collision occurred between 57 Ma and 47 Ma. A review of published data provides further evidence of &50 Ma magmatism and Cu-Au-Mo porphyry deposit mineralisation north of the Himalayan range. These rocks would be interpreted to have formed in an arc setting in other parts of the world. However, most workers consider that &50 Ma magmatism and porphyry deposit formation in the Himalaya is unrelated to subduction of oceanic lithosphere and are instead considered to form owing to 'post-collisional' processes. We argue that these interpretations have been based on an established model where the India-Eurasia collision is deemed to occur at ca 50 Ma, instead of using the data to question and refine existing tectonic models. © 2012 Copyright Taylor and Francis Group, LLC.