Subsurface fault geometry is one of the most difficult parameters to constrain in geophysical studies, and yet an accurate model of the geometry is critical for unlocking our understanding of the physical properties and behaviour of faults. In the Himalaya, a combination of dense geologic, microseismic, and geodetic observations allow us to investigate this effect throughout the seismic cycle. Our results show that the ramp-flat geometry of the Main Himalayan Thrust (MHT), long proposed from structural observations, is the key geometric feature of the MHT and exerts a first-order control on the fault’s interseismic and coseismic behaviour. The motivation for this study is the 2015 Mw7.8 Gorkha, Nepal earthquake, where geologic reconstructions of the MHT geometry suggest that the rupture occurred on a flat bounded on all sides by ramps. During the interseismic period, we construct a simplified model of the fault coupling and find that the locked-to-creeping transition is in close agreement with geologic reconstructions of the midcrustal ramp location throughout Nepal. The inferred width of coupling varies significantly along strike, from 100-110 km in central Nepal to 70-90 km in eastern and western Nepal. In western Nepal, microseismic ang geologic evidence suggest two possible mid-crustal ramp locations; the geodetic data are in better agreement with the southern location. This reduces the width of the inferred locked zone by up to 30% compared with earlier models that assumed the northern location, and suggests that an active duplexing process in this area may be nearly complete. We find that the data also show a systematic decrease in the total rate of convergence along strike, from 18-20 mm/yr in western Nepal to just 13-14 mm/yr in central and eastern Nepal. This reduced rate has important implications for models of earthquake recurrence rates, as well as for reconstructions of total shortening across structural cross sections.