<p>In the present study, the Expert Team on Climate Change Detection and Indices (ETCCDI) for extreme rainfall over India in the Coupled Model Intercomparison Project Phase-6 (CMIP6) historical simulations and future projections are assessed before and after downscaling and bias correction (DBC). The relative performance of the Multi-model mean (MMM) and individual models in representing the frequency and intensity of extreme rainfall events over India and their significant improvement after DBC, is reported. The representation of RX1Day, RX5Day, R99p, R95p, R20mm, and R10mm over India has improved by 83.18%, 80.74%, 94.36%, 83.85%, 64.37%, and 33.25%, respectively in AD-MMM (MMM after DBC) compared to BD-MMM (MMM before DBC). Higher extreme indices, such as RX1Day and R99p, are well captured in the DBC historical product, an important information for choosing the right index for estimating the CMIP6 model projected extremes. However, milder extreme indices, including RX5Day and R95p, and lower extreme indices, like R20mm and R10mm, exhibit limited skill in representing the characteristics of extreme rainfall. The projections indicate that under SSP245 (SSP585), RX1day is expected to increase by 38% (41.65%) in the near future and by 48.53% (62.07%) in the far future. Similarly, extremes based on R99p are projected to increase by 33.12% (37.57%) in the near future and 44.32% (59.33%) in the far future over the Indian region. Both indices indicate a 1.61-fold increase in extreme rainfall over the Indian subcontinent in the far future under the SSP585 scenario compared to the historical period. Particularly, in the near future, the projected increase is expected to be more pronounced over northwest India, while in the far future, the central, northwest and northeast region is projected for most significant rise. In short, this study highlights the potential of CMIP6 models in capturing extreme rainfall indices and their future projections.</p>