Mandatory water use restrictions have become a common feature of the urban water management landscape in countries like Australia. Water restrictions limit how water can be used and their impacts have often been enumerated by using stated preference techniques, like contingent valuation. Most interest in these studies emerged in times of drought, when the severity of restrictions and their deployment had increased and water managers contemplate supply augmentation measures. A question thus arises as to whether the same estimates can be legitimately deployed to water supply projects undertaken when water is more plentiful. This study sheds light on the impact on estimates of willingness to pay when the climatic backdrop to a contingent valuation experiment is altered. We report the results of a comparison between two surveys, undertaken in 2008 and 2012, using a common multiple-bounded discrete choice contingent valuation design, administered across six cities in Australia, covering metropolitan and regional settings. Using a finite mixture, scaled ordered probit model we investigate changes over time in willingness to pay by city, and also causes of individual heterogeneity in willingness to pay. We find that willingness to pay estimates significantly change over time in most regional centres but this is not the case for the major cities of Sydney and Melbourne, once changes in housing prices are included in the analysis.