We use teleseismic data from the Jammu and Kashmir Seismological NETwork, to perform P-wave receiver function spatial and common-conversion-point (CCP) stacks, and joint inversion with Rayleigh-wave group-velocity dispersion, to construct 3D Vs model of the Jammu and Kashmir (J&K) Himalaya. 2D-CCP and Vs profiles reveal increasing crustal thickness from the foreland-to-hinterland, and an under-thrust Indian crust beneath J&K. The Moho positive impedance-contrast boundary is at ∼45 km depth beneath Sub-Himalaya and deepens to ∼70 km beneath Higher-to-Tethyan Himalaya, with an overall gentle NE dip. The Main Himalayan Thurst (MHT) forms a low velocity layer (LVL) with negative impedance contrast, and has a flat–ramp geometry. The flat segment is beneath Sub-to-Lesser Himalaya at 6–10 km depth, and dips ∼4◦. The mid-crustal (frontal) ramp is beneath Kishtwar Higher-Himalaya and Zanskar Ranges at 10–16 km depth, and dips ∼13–17◦. Significant along-arc variation in crustal structure is observed between east (Kishtwar) and west (Kashmir Valley) segments. Beneath the Kishtwar Window we image a Lesser Himalayan duplex (LHD) bound between MHT sole-thrust and MCT roof-thrust. LHD horses dip at high angle to the bounding structures and are illuminated by moderate seismicity. Beneath the Pir-Panjal Ranges and Kashmir Valley, the underthrust crust is ∼10 km thicker, has higher crustal Vs , and a shallower flat MHT at ∼10 km depth. The westward shallowing of the MHT occurs through a lateral ramp beneath Kishtwar Himalaya. Aftershocks of the 2013 Kishtwar earthquake concentrate on the MHT frontal and lateral ramp intersection, and possibly marks the down-dip locked-to-creep transition.