Different Life Cycle Impact Assessment (LCIA) methods have been developed over the past twenty years. While the use of a single score indicator may, on the one hand, ease the decision making it may also induce a loss of information. On the other hand, the use of a large set of environmental indicators may increase the difficulty of decision making due to the high number of parameters. In this paper, a statistical methodology is used to identify a simplified set of environmental indicators. This methodology applies Principal Component Analysis (PCA) to five LCIA methods (CML’01, Eco-indicator99, IMPACT 2002+, ecological scarcity, EPD) using a building products database. For each method, the results show that only 4–6 dimensions are sufficient to explain at least 90–95% of the variance for each set of indicators. They refer to the following environmental themes: fossil-fuel consumption, ecotoxicity, ionising radiation, land use and mineral resources. The use of a selected set of five LCI flows is found to be as relevant as current LCIA methods from a statistical point of view. Furthermore, PCA applied to a building LCA case study shows similar trends as for the database results. Further studies are now needed to confirm these findings for going towards simplified structural relationships among LCIA indicators for building materials.