Purpose: Number of extreme rainfall is increasing in Japan and these damages are also become obviously. Even we had huge damages in local areas in 2015 and 2016. Japan has tackled with flood and inundation reduction since ancient times but the potential protection level is decreasing now because of climate change. Quantitative evaluation of adaptation for flooding is requested and cost value is useful for discussion of policy decisions. Numerical flood simulation and some GCMs with scenarios can evaluate adaptation effects. Also spatial information showing vulnerable areas is important for policy making. Therefore the distribution of damage cost was made in this study and we discussed the adaption in spatial and temporal. Methodology: 2D non-uniform flow is modeled to obtain the inundation depth and duration, which are used for calculation of flood damage in each landuse according to the economic manual for flood protection (MLIT, Japan). Extreme rainfall as a distribution map is estimated from GEV distribution function using 30 years data, interpolated observed data into a whole Japan, and input these to the model. 4 GCMs and 3 RCP scenarios provide future rainfall, which is used for downscaling and bias correction. Present flood damage is calculated by the difference between damage of GCM input rainfall and 50 years return period rainfall, which is supposed as average flood protection level in whole Japan. Adaptation effect is evaluated as the damage to flood protection of 70 years return period rainfall, which is 20 years return period protection is higher than current flood protection level. Summary: In the case of current flood protection level, future flood damage costs are much different for each GCM with scenario which has double value on the estimation. Roughly the damage cost will be 2 to 3 times in around 2100 comparing with current situation. In the case of adapted flood protection level by increase of 20 years return period for peak discharge, some of all damage costs by all GCMs will be negative to current damage and the average future damage cost with the adaptation is similar to current damage cost that the shift of damage cost in the future is similar to the investment of 20 years return period protection. Even now, Japanese government is continuing flood control but the rate of damage for the future is higher than the increase of flood control by hardware such as dams, dikes and retention ponds. It would be difficult to raise the level for 20 years flood protection in whole Japan until 2100, therefore, all manner of countermeasures are necessary for flood adaptation such as early warning system, evacuation drills and others.