<p>Precipitation extremes are critical for Australia’s water and agriculture, yet regional climate models contain systematic biases that distort rainfall and drought estimates. This study evaluates the added value of bias correction applied to high-resolution CCAM-ACS simulations, with a particular focus on precipitation extremes, drought characteristics, and their implications for wheat exposure. A pseudo-future split-sample experiment (1955–1984 calibration; 1985–2014 validation) is used to assess the robustness of Quantile Delta Mapping (QDM), applied at a grid-by-grid and monthly scale to seven CMIP6-driven CCAM-ACS models, with projections for 2030–2059 under SSP1-2.6 and SSP3-7.0. The raw simulations exhibit pronounced spatial biases, particularly over the monsoonal north and temperate southern regions. Bias correction substantially reduces mean precipitation errors, lowering the overall RMSE from approximately 22 to 5 mm month⁻¹, and improves the statistical representation of precipitation distributions. Improvements in extremes are assessed separately using monthly distributional metrics (e.g., Q10 and Q90), showing enhanced representation of both dry and wet tails. Improvements are evident across most regions, although residual biases remain, especially in areas dominated by convective processes. Future projections indicate under SSP3-7.0 indicate drought-affected areas expand by 5–10%, with intensified severe droughts in southern Australia and shorter, sharper dry spells in the north. These changes are associated with increased drought severity and variability, with implications for agricultural productivity. Changes in event frequency remain modest, with dry-event increases of up to ~2 percentage points in southern and interior regions and slight decreases in wet-event frequency in southern areas, indicating a gradual strengthening of the north–south hydroclimatic gradient rather than a substantial intensification of extremes. Wheat-cropping area also shows strong spatial heterogeneity: Murray-B and Southern-S remain relatively stable, whereas Range-L, North-M and East-C show marked increases in zero or near-zero wheat-area spatial units, reaching 49.0–51.7% in Range-L and 75.5% in North-M under SSP3-7.0. Nationally, near-zero wheat-area spatial units increase from 0.1% historically to 7.6% under SSP1-2.6 and 9.8% under SSP3-7.0. These findings demonstrate that bias correction enhances the statistical consistency of climate simulations and supports more robust assessment of drought and climate–agriculture interactions.</p> Graphical abstract <p>The graphical abstract illustrates the added value of bias correction for high-resolution CCAM-ACS simulations and its implications for precipitation extremes, drought risk, and wheat production across Australia. Raw model outputs exhibit systematic precipitation biases, with wet biases of about +45 mm month⁻¹ in northern regions and dry biases of −30 mm month⁻¹ in southern Australia. These distortions extend to the upper rainfall tail, where 90th-percentile precipitation (Q90) is overestimated, resulting in a biased statistical representation of extreme rainfall and apparent increases in drought persistence metrics. It then visually demonstrates the application of Quantile Delta Mapping (QDM) within a pseudo-future split-sample framework (1955–1984 calibration; 1985–2014 validation), highlighting substantial reductions in mean bias (70–85%) and RMSE (22 to 5 mm month⁻¹), along with added value exceeding +10 mm month⁻¹ across most regions. Projections for 2030–2059 show that SSP3-7.0 intensifies drought exposure across key southern and interior agricultural regions, with drought-affected areas expanding by approximately 5–10% and severe to extreme droughts exceeding 20–30% of regional area in some peak years. Wheat-cropping vulnerability is expressed mainly through spatial contraction and fragmentation, with zero or near-zero wheat-area spatial units reaching 49.0–51.7% in Range-L and 75.5% in North-M, while Murray-B and Southern-S remain comparatively stable. Overall, the graphical abstract integrates climate modelling, bias correction, drought dynamics, and spatially explicit wheat-cropping exposure into an accessible visual summary for regional adaptation planning.</p>