We have only rudimentary understanding of the complex and pervasive connections between water and energy in cities. As water security now threatens energy and economic security, this is a major omission. Understanding the water-energy nexus is necessary if we want to contribute to solving water and energy issues simultaneously; if we want to stop moving problems from one resource dimension to another. This is particularly relevant in the Australian context where energy use for water supplies is forecast to rapidly escalate, growing around 300% from 2007 levels, by 2030. This paper presents a literature review with an aim of characterising the research to date with a particular focus on cities, the major centres of consumption and growth. It systematically analyses a wide range of papers and summarises the diverse objectives, dimensions, and scale of the research to-date together with knowledge gaps. There are many major gaps. These include energy use associated with water in industrial and commercial operations as well as socio-political perspectives. A major gap is the lack of a unifying theoretical framework and consistent methodology for analysis. This is considered a prerequisite for quantitative trans-city comparisons.

This paper reviews current knowledge about water footprints (WFs) and the role of input–output techniques. We first provide an overview of the prevailing ‘bottom-up’, process-based methods and their strengths and limitations. This overview leads to discussion of the benefits of combining process-based water footprints with information from input–output techniques. The central theme and proposition is that environmental multi-region input–output analysis (E-MRIO) has a powerful capacity to establish the geography of embodied water, and to complement process-based approaches to WF by expanding their supply-chain coverage. Combining process and input–output information provides valuable information for a diverse set of water planning and water policy objectives. A comprehensive and systematic outline of potential policy applications of E-MRIO (and process analysis methods) is presented.