The Energy Efficiency First (EE1st) principle aims to consider and prioritise demand-side resources (end-use energy efficiency, demand response, etc.) whenever they deliver more value than supply-side resources (generation, networks, storage, etc.) in achieving the same objectives. This study set out to provide quantitative evidence on EE1st by investigating the level of end-use energy efficiency measures in the building sector that would provide the greatest benefit for the European Union (EU) in transitioning to net-zero greenhouse gas (GHG) emissions by the year 2050. By using a suite of energy system models, three scenarios are investigated (LOWEFF, MEDIUMEFF, HIGHEFF), each of them reaching the net-zero target. The scenarios differ regarding the ambition level for building retrofits, energy-efficient appliances and other energy efficiency measures in residential and non-residential buildings. These differences, in turn, affect the deployment of energy generators and networks for electricity, district heating, and hydrogen. As such, this analysis helps determine the extent to which society is better off – in pure monetary terms – if end-use energy efficiency in buildings was prioritized over generators, networks and storage facilities, in line the EE1st principle. The findings suggest that energy efficiency in buildings is critical to achieve net-zero GHG emissions by 2050. The LOWEFF scenario (-21.1% reduction in final energy demand for buildings in 2050 vs. 2020 levels) represents the conservative lower end of reasonable ambition levels for end-use energy efficiency in buildings with a view to net-zero emissions. This level of ambition is significantly above the business-as-usual pathway of the EU Reference Scenario (-10.4% final energy demand in 2050 vs. 2020) (Capros et al. 2021). According to the central indicator of energy system cost, the more ambitious scenarios MEDIUMEFF (-30.2% final energy demand in 2050 vs. 2020) and HIGHEFF (-35.5% final energy demand in 2050 vs. 2020) are generally not cost-effective in comparison to LOWEFF. However, there is ample reason to support these ambition levels beyond the LOWEFF scenario. For one thing, the differences in energy system cost are small in magnitude – e.g. the additional annual cost in HIGHEFF vs. LOWEFF corresponds to less than 0.03% of the EU’s gross domestic product. For another, this study did not anticipate the recent spike in energy prices as of 2021–2022, which would justify higher ambition levels for energy efficiency (Eichhammer 2022). The same applies to the inclusion of indoor comfort gains, reduced air pollution and other multiple impacts. As demonstrated in a follow-up report, their consideration significantly enhances the attractiveness of energy efficiency and thus provides further support for the EE1st principle. In practice, the scenarios set out in this study require an ambitious package of planning and policy instruments. Another branch of reports in the ENEFIRST project provides a detailed account of policy design options for EE1st.