The pattern of fluvial incision across the Himalayas of central Nepal is estimated from the distribution of Holocene and Pleistocene terraces and from the geometry of modern channels along major rivers draining across the range. The terraces provide good constraints on incision rates across the Himalayan frontal folds (Sub-Himalaya or Siwaliks Hills) where rivers are forced to cut down into rising anticlines and have abandoned numerous strath terraces. Farther north and upstream, in the Lesser Himalaya, prominent fill terraces were deposited, probably during the late Pleistocene, and were subsequently incised. The amount of bedrock incision beneath the fill deposits is generally small, suggesting a slow rate of fluvial incision in the Lesser Himalaya. The terrace record is lost in the high range where the rivers are cutting steep gorges. To complement the terrace study, fluvial incision was also estimated from the modern channel geometries using an estimate of the shear stress exerted by the flowing water at the bottom of the channel as a proxy for river incision rate. This approach allows quantification of the effect of variations in channel slope, width, and discharge on the incision rate of a river; the determination of incision rates requires an additional lithological calibration. The two approaches are shown to yield consistent results when applied to the same reach or if incision profiles along nearby parallel reaches are compared. In the Sub-Himalaya, river incision is rapid, with values up to 10--15 mm/yr. It does not exceed a few millimeters per year in the Lesser Himalaya, and rises abruptly at the front of the high range to reach values of ~4--8 mm/yr within a 50-km-wide zone that coincides with the position of the highest Himalayan peaks. Sediment yield derived from the measurement of suspended load in Himalayan rivers suggests that fluvial incision drives hillslope denudation of the landscape at the scale of the whole range. The observed pattern of erosion is found to closely mimic uplift as predicted by a mechanical model taking into account erosion and slip along the flat-ramp-flat geometry of the Main Himalayan Thrust fault. The morphology of the range reflects a dynamic equilibrium between present-day tectonics and surface processes. The sharp relief together with the high uplift rates in the Higher Himalaya reflects thrusting over the midcrustal ramp rather than the isostatic response to reincision of the Tibetan Plateau driven by late Cenozoic climate change, or late Miocene reactivation of the Main Central Thrust. ¿ 2001 American Geophysical Union