Purpose <p>This paper compares the environmental impact of the European residential stock as calculated with two LCA-based building stock models (Consumption Footprint – BoP Housing and PULSE-EU). The goal is to identify the purpose for which each model is best suited and to reflect on the need to balance modelling feasibility with accuracy, the uncertainties inherent to impact estimation at a large scale, and how they can affect policy support efforts.</p> Methods <p>Whole life impacts are assessed first, followed by a separate analysis of operational and embodied impacts. To allow for an accurate comparison, the models are harmonised, leading to a partial scope adjustments, the selection of the same baseline year (2020) and the calculation of impacts using the indicators of the EN15804 + A2 standard. Discrepancies and commonalities are identified and their link to the characteristics of each model (such as granularity, archetype characterisation, and the approach used to model individual life cycle stages) is investigated. Crucial differences are identified in terms of modelling choices, particularly as it concerns background and foreground data selection, use of bottom-up and top-down data, upscaling parameters and the approach to emission timing.</p> Results and discussion <p>The results of the two models fall within the same order of magnitude for most indicators, though the percentage variation varies significantly. Operational impacts are higher in BoP Housing, and embodied impacts in PULSE-EU. These discrepancies are linked to different methodological choices. The study highlights the challenges associated with accurately estimating the environmental impact of a complex system such as the EU residential sector. The analysis reveals that both models have their strengths and weaknesses. Their coexistence can become instrumental in recognising the uncertainties inherent to this type of analysis, and provide a more nuanced interpretation of its results. BoP Housing, with its yearly updates, can track the evolution of the building sector over time; PULSE-EU, with its higher granularity, is well-suited to scenario analysis.</p> Conclusions <p>In any building stock model, it is necessary to find an equilibrium between level of granularity and model flexibility. To reduce the uncertainties inherent to the models, it is important to promote data collection and accessibility within the EU. These uncertainties, however, can also be used to achieve a deeper understanding of the results obtained, leading to a more nuanced interpretation and a critical analysis of the way they are used in policy support.</p>