I discuss the danger of earthquakes occurring directly beneath the metropolitan area from temporal and geographical viewpoints. Temporally, large (M>7) earthquakes in Kanto occur 70 years before and a few years after great interplate Kanto earthquakes. The recurrence times of such great earthquakes are more than 220 years. Because 80 years have passed since the last one, at least60 years remain before reaching the active period. It is not legitimate to calculate the probability of large earthquakes occurring in the near future, using the rate of occurrence during the active period before a great earthquake. From a geographical viewpoint, S. Kanto is located in the outer zone south of the Median Tectonic Line, where few active faults are distributed. However, in S. Kanto, exceptionally, the Tachikawa fault and the 1855 Ansei-Edo earthquake are located in the outer zone. This zone is specified by a low-velocity zone in the mantle wedge of the upper plate. Dehydration from the subducting slab may weaken the upper plate in this zone, producing anomalous intraplate earthquakes. The upper crust above this low-velocity area should be marked especially as an area having a potential for large earthquakes in the future. The probability of M7 earthquakes being generated at the interface between the Philippine Sea and the upper plates is small. Temporally, the danger of large earthquakes occurring in the near future beneath the metropolitan area does not seem to be large.
