Despite significant efforts to quantify the mutual inter-dependence of the water and energy sectors, global energy for water (EFW) remains poorly understood, resulting in biases in energy accounting that directly affect water and energy management and policy. We firstly evaluate the global energy consumption for direct water use from 1973 to 2012 with sectoral, regional and process-level details. Over the 40-year period, we detected multiple shifts in EFW by county and region. For example, we find that India, the Middle East and China have surpassed the United States as the three largest consumers of EFW since 2003, mostly because of rapid growth in groundwater-based irrigation, desalination, and industrial and municipal water use, respectively. Globally, EFW accounts for 1–3% of total primary energy consumption in 2010, of which 52% is surface water, 36% is groundwater, and 12% is non-fresh water. The sectoral allocation of EFW includes municipal (45%), industrial (29%), and agricultural use (26%), and process-level contributions are from source/conveyance (41%), water purification (19%), water distribution (13%) and wastewater treatment (22%). Our evaluation suggests that the EFW may increase in importance in the future due to growth in population and income, and depletion of surface and shallow aquifer water resources in water-scarce regions. We are incorporating this element into an integrated assessment model (IAM) and linking it back to energy balance within that IAM. By doing this, we will then explore the impacts of EFW on the global energy market (e.g., changes in the share of groundwater use and desalination), and the uncertainty of future EFW under different shared social pathway (SSP) and representative concentration pathway (RCP) scenarios, and consequences on the emission of greenhouse gases as well. We expect these EFW induced impacts will be considerable, and will then have significant implications for adaptive management and policy making.