Geodetic estimates of crustal strain rates in the Japan islands were an order of magnitude larger than geological/geomorphological estimates, which has been an unresolved problem called the strain rate paradox. Ikeda (1996) postulated that geodetic strain mainly reflects elastic strain accumulation due to interactions at plate boundaries. This hypothesis was proven to be correct by the occurrence of the 2011 MW 9.0 Tohoku-oki earthquake. Confusion between elastic strain and inelastic strain was the cause of the paradox. Significant postseismic deformation observed after the Tohoku-oki earthquake made it possible to distinguish the inelastic contribution from the geodetically observed crustal strain through a comparison with the pre-seismic strain rate pattern, which promoted a better understanding of inelastic deformation in the Japan islands. On the other hand, migration of localized deformation and temporal changes of strain rate are identified over a geological time scale, implying that it is essential to carefully review the methods and the uncertainties of geological/geomorphological strain rates. An integrated understanding of crustal deformation in the Japan islands is being advanced through detailed investigations of crustal strain rates on variable temporal and spatial scales.