Accurate projections of water demand extremes are essential to all aspects of the planning process (e.g., defining target growth and securing financing). The role of climate on water demand variability is pronounced and will likely increase in the future as a result of climate change. Therefore, a flexible approach that can incorporate climate change scenarios is needed to project water demand extremes. In this research, an integrated approach with two components was developed. First, extreme value models based on climate attributes were created for both water demand maximums and water demand over a specified threshold; these models were then fitted to the historic water demand and climate data. Second, ensembles of future weather sequences conditioned on climate change projections were generated using a stochastic weather generator. Together these two components provided scenarios of water demand extremes under various climate projections. The city of Aurora, Colo., was used as a case study to demonstrate the utility of this approach.