The 2015 M w 7.8 Gorkha earthquake in Nepal and the M w 7.3 aftershock triggered at least 22,914 landslides that each had areas ≥500 m 2 and lengths and widths ≥20 m. Amongst these landslides, 2,059 had areas >10,000 m 2. Analyses of the landslide geometry, using length (L), width (W), height (H, from the crown to toe), reach angle (arctan value of H/L), and aspect ratio (L/W), show that most of the landslides have aspect ratios of 1.6-3.6 and reach angles of 35-45°. The fitting relationship between H and L is H 0.87L − 11.11. The steep topography is likely the main factor that controls the landslide runout and planar shape. The landslides are divided into 3 geomorphometric categories using the aspect ratio: LS1 (L/W ≤ 2); LS2 (2 < L/W ≤ 4); and LS3 (L/W > 4). Statistical analyses of these categories with the control factors show that the landslide distribution does not relate to the three largescale geologic faults that traverse the region, roads, accumulative precipitation before the earthquakes, and the small earthquakes that occurred during the 2012-2015 premonsoons in the study area. The 3 landslide categories are sensitive to similar conditions related to curvature, slope position, lithology, and peak ground acceleration. In contrast, the effects of elevation, slope angle, slope aspect, and streams on landslide distribution differ. Moreover, massive landslides (with areas >10,000 m 2) are more likely to occur on the steeper hill slopes that in the higher elevation settings, which provide more substantial gravitational potential energy and long-runout space. As landslides with different geomorphometric shapes have various susceptible conditions, examining the landslide distribution based on their geometric characteristics provides a new way to study the landslide extent and mechanism.