The quantile mapping (QM) bias correction algorithm was applied to bias-correcting the daily rainfall over the Malaysia based on the APHRODITE gridded observed dataset. The quality of the bias correction in reducing the biases of regional climate models in frequency (FREQ), mean intensity (SDII) and 90th percentile (Q90) of the daily rainfall was assessed. The result shows that QM performs well in correcting the rainfall statistics. The QM bias correction method was then applied to seven Regional Climate Models (RCMs) projection output from the CORDEX Southeast Asia (CORDEX-SEA) dataset, to obtain high resolution rainfall change scenarios over the Malaysia, based on the IPCC’s representative concentration pathways (RCP) 4.5 and 8.5 scenarios. The ensemble mean of seven RCMs suggests that the future projection expected a decrease in the rainfall frequencies over the Malaysia toward the 21st century for both RCP scenarios. Meanwhile, the rainfall intensity and extremes over the Peninsular Malaysia (East Malaysia) is likely to decrease (increase) especially during the winter (summer and autumn) months by the end of the century. A remarkable decrement of rainfall frequency and extremes was found in western Peninsular Malaysia, with changes of around −4 to −8% with respect to the historical period driven by the RCP4.5 and RCP8.5 scenario respectively.