Purpose : Life cycle assessment can support decisions for improving the environmental performance of construction products. However, the amount of data required for developing life cycle inventories limits the adoption of LCA. This work associates the interpretation of the impact results of construction products at the unit process level with a quantitative definition for the foreground and background system, for guiding primary data collection towards foreground processes that can be affected by decision-makers in the construction sector. Methods : A set of construction products commonly used in Brazil is selected, and their cradle-to-gate life cycle inventories are modeled using the ecoinvent database (version 2). Life cycle impact assessment is performed using the ReCiPe Midpoint Hierarchist method. The contribution of each process during the life cycle of construction products for each impact category is quantified. These processes are associated with economic sectors, which are classified as belonging to the foreground or background system from the perspective of the construction industry. Foreground sectors are those controlled or influenced by the construction sector and are defined based on the production share consumed by the construction value chain. The elementary flows defining each impact category are also identified. Results and discussion : Foreground processes show significant contributions to most impact results of construction products. Global warming, fine particulate matter formation, ozone formation, acidification, human carcinogenic toxicity, and terrestrial ecotoxicity are mainly caused by direct emissions and fossil fuel combustion in manufacturing processes. Land occupation for production activities contributes to land use change, while the consumption of fuels, raw materials, and water causes fossil and mineral resource scarcity and water consumption respectively. Freshwater and marine ecotoxicities and human non-carcinogenic toxicity have foreground contributions only for steel and copper products due to emissions from the landfilling of mining tails. Ionizing radiation and stratospheric ozone depletion are mostly driven by background processes. A reduced group of elementary flows covers a big share of the environmental impacts of most construction products. Conclusions : The results indicate priorities for life cycle inventory primary data collection of construction products, by focusing on foreground processes and the corresponding elementary flows that cause many of the potential embodied impacts of construction. Increasing the availability of primary data for these processes improves the reliability of LCA-based decisions in the construction sector, especially in countries which still lack local LCI databases.