Seismic risk for the world's major cities: Estimates verified T. L. Kronrod and A. K. Nekrasova Abstract We compare theoretical assessments of seismic risk for the world's major cities derived by Keilis-Borok et al. (1985) for the period 1971--2000 and observed earthquake damage during the past 25 years. Two risks of large earthquakes are considered, the number of cities, $N_t$, and the number of their residents $D_t$ in the zone of catastrophic shaking. We have good agreement between the number of cities hit by earthquakes and the number of actual shakings for the 25-year period ($6.7\pm3.2$ were predicted, while 6 or 7 were damaged, there being no reliable information for Kyoto). No reliable information on the population in the zone of strong shaking is available, but the reliability of the prediction of $D_t$ is indirectly confirmed by comparing the population as predicted for 1990 and the actual population. It is shown that the worldwide agglomeration of cities and megalopolis development affect risk to urban populations and economies in a significant manner. Back to Computational Seismology, Vol. 4.