Recently frequency and intensity of natural disasters have been increasing due to severe climate change. Because most damages of natural disaster over the Korean Peninsula have been related to extreme precipitation events, it is important to predict future changes of extreme precipitation phenomena caused by climate changes. This study analyzed changes in extreme precipitation indices, which were calculated from detailed climate change scenarios produced by multi regional climate models. 6 core precipitation indices made by STARDEX, which is one of the projects in EU, were chosen. Changes of STARDEX indices of precipitation were analyzed based on the differences between present 25-year (1981-2005) HISTORIDCAL data and future 25-year (2076-2100) RCP 8.5 data by 5 RCMs (HadGEM3-RA, RegCM4, SNURCM, WRF, GRIMs). Gridded precipitation data from RCMs were bilinearly interpolated into 230 cities in South Korea, and the STARDEX indices were calculated for each cities. As a result, 4 of 5 RCMs predicted increasing intensity of precipitation, especially for southern part of South Korea. On the other hands, RCMs predicted increasing maximum consecutive dry day in central region of South Korea. Most of extreme indices changes were related to changes in synoptic conditions (e.g., monsoonal wind, upper-level jet, and subtropical high) and convective instability. It is expected that the results of this study (i.e., future indices changes) can be fruitful references for policy making of disaster management.