Spatial variability in precipitation has received little attention in the study of connections between climate, erosion, and tectonics. However, long-term precipitation patterns show large variations over spatial scales of ∼10 km and are strongly controlled by topography. We use precipitation rate estimates from Tropical Rainfall Measuring Mission (TRMM) satellite radar data to approximate annual precipitation over the Himalaya at a spatial resolution of 10 km. The resulting precipitation pattern shows gradients across the range, from east to west along the range, and fivefold differences between major valleys and their adjacent ridges. Basin-wide average precipitation estimates correlate well with available measured mean runoff for Himalayan rivers. Estimated errors of 15%–50% in TRMM-derived annual precipitation are much smaller than the spatial variability in predicted totals across the study area. A simple model of orographic precipitation predicts a positive relationship between precipitation and two topographically derived factors: the saturation vapor pressure at the surface and this pressure times the slope. This model captures significant features of the pattern of precipitation, including the gradient across the range and the ridge-valley difference, but fails to predict the east-west gradient and the highest totals. Model results indicate that the spatial pattern of precipitation is strongly related to topography and therefore must co-evolve with the topography, and suggest that our model may be useful for investigation of the relationships among the coupled climate-erosion-tectonic system.