Quantitative estimates of casualties calculated in models of future earthquakes are essential for realistic preparation for earthquake disasters and for risk reduction. We estimate the number of casualties and strongly affected people (living in the area of intensities MMI [modified Mercalli intensity] = VI+) in case the 1555 and the 1505 Himalayan earthquakes should repeat with similar magnitudes and locations. The source areas, magnitudes, and depths of our scenario earthquakes are derived from parameters given in the literature. Before calculating scenario losses, we validate the loss estimation software QLARM for the area by matching observed intensities and fatality counts, given the source parameters of five historic ruptures: the 2011 M 6.9 Sikkim, the 2005 M 7.6 Kashmir, the 1999 M 6.6 Chamoli, the 1991 M 6.8 Uttarkashi, and the 1905 M 7.8 Kangra earthquakes. Based on the quality of matches, we use somewhat stronger than average attenuation for the scenario loss estimates. Choices of parameters that keep our loss estimates low are hypocentral depths deeper and rupture lengths shorter than values that could also be assumed. For a repeat of the 1555 earthquake, we offer two scenario estimates with M 7.6 and 8. Assuming M 8.0, we estimate 221,000 fatalities, 884,000 injured, and 94 million strongly affected people. For a repeat of the 1505 earthquake, our two scenarios assume M 8.7, and the depths are assumed as 15 and 20 km, respectively. Using a depth of 15 km, we estimate that an earthquake similar to that of 1505 may result in 599,000 fatalities, 1.1 million injured, and 78 million strongly affected people. Our order of magnitude estimates are more quantitative but agree in general with two other loss estimates for Himalayan earthquakes based on different scenarios and methods. To cope with disasters of this scale, great efforts and careful planning are required.