We review earthquake observation and strong motion seismology in Japan over the three decades starting in 1975. Preceding the 1995 Kobe earthquake, earthquake prediction research programs played an important role in earthquake observation research. The devastating damage from this earthquake, however, forced a change in emphasis from empirical shortterm prediction to long-term forecasting of earthquakes and the prediction of strong ground motion. Nationwide observation networks were set up, and progress in strong motion seismology was applied to projects of national seismic hazard maps. The next disastrous earthquake may even force their reexamination in the near future.