Current general circulation models (GCMs) is limited by the fact that the spatial and temporal resolutions are insufficient to provide the details required for assessing changes in extreme rainfall events. To resolve processes and features at the right scale, regional climate models (RCMs) forced with GCM data, are commonly used. Although this provides much higher resolution than the global model, however, their application is hindered by systematic biases in large-scale circulation patterns from driving GCM data which can be amplified further by the RCM. To deal with these considerable biases, recent studies have suggested the bias correction of the RCM input boundary conditions. This study focuses on the impact of bias corrections on the lateral boundary conditions on precipitation extremes. Three bias correction methods are used including mean, mean and variance, and nested bias correction (NBC) that corrects for lag-1 autocorrelations. The European Center for Medium-Range Weather Forecast’s (ECMWF) ERA-Interim (ERA-I) reanalysis model is used here as “perfect boundary conditions” for bias correction. Weather Research and Forecasting model (WRF), a next generation mesoscale RCM model, is used in this study. The downscaling is performed over the Australasian Coordinated Regional Climate Downscaling Experiment (CORDEX) domain. To quantify the impact of bias correction on the WRF output, we have used two quantitative measures: root-mean-square errors (RMSE) and bias. The results are then evaluated comparing ERA-I-driven WRF simulation and the bias correction cases. The results show that bias correction on the lateral boundary conditions produces a noticeable improvement in precipitation extremes.