We analyze geomorphic evidence of recent crustal deformation in the sub-Himalaya of central Nepal, south of the Kathmandu Basin. The Main Frontal Thrust fault (MFT), which marks the southern edge of the sub-Himalayan fold belt, is the only active structure in that area. Active fault bend folding at the MFT is quantified from structural geology and fluvial terraces along the Bagmati and Bakeya Rivers. Two major and two minor strath terraces are recognized and dated to be 9.2, 2.2, and 6.2, 3.7 calibrated (cal) kyr old, respectively. Rock uplift of up to 1.5 cm/yr is derived from river incision, accounting for sedimentation in the Gangetic plain and channel geometry changes. Rock uplift profiles are found to correlate with bedding dip angles, as expected in fault bend folding. It implies that thrusting along the MFT has absorbed 21¿1.5 mm/yr of N-S shortening on average over the Holocene period. The ¿1.5 mm/yr defines the 68% confidence interval and accounts for uncertainties in age, elevation measurements, initial geometry of the deformed terraces, and seismic cycle. At the longitude of Kathamandu, localized thrusting along the Main Frontal Thrust fault must absorb most of the shortening across the Himalaya. By contrast, microseismicity and geodetic monitoring over the last decade suggest that interscismic strain is accumulating beneath the High Himalaya, 50--100 km north of the active fold zone, where the Main Himalayan Thrust (MHT) fault roots into a ductile d¿collement beneath southern Tibet. In the interseismic period the MHT is locked, and elastic deformation accumulates until being released by large (Mw>8) earthquakes. These earthquakes break the MHT up to the near surface at the front of the Himalayan foothills and result in incremental activation of the MFT. ¿ 2000 American Geophysical Union