Severe damage from the 1995 Hyogoken-Nanbu earthquake gave us a lesson that a cooperation of research of active fault and strong ground motion is very important for the prediction of strong ground motion from inland shallow earthquake. That is an ace of the scientific world for the mitigation of earthquake damage. In this paper we take up two subjects which are important to put the strong motion prediction to the practical use. One is to estimate the magnitude of a future big earthquake from the active fault data and the other is to estimate the upper level of strong ground motions caused by blind faults. Expectations for active fault research to realize a useful strong motion prediction are described.