Abstract The built environment stocks such as buildings and infrastructures are key to human development: they provide the fundamental physical settings that the provision of basic human needs such as food, shelter, and transport rely on, but also contribute to anthropogenic greenhouse gas (GHG) emissions throughout their construction, operation, and end-of-life management phases. These stocks usually exist in societies for relatively long time, from years to over a century, therefore their dynamics have long term impacts on human development and emission growth. Several recent studies, including the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC), have discussed the lock-in effects of infrastructure stocks on emission pathways. However, there is still a lack of quantitative analysis and evidence to support this claim. Here, based on an empirical regression model and a new dataset that determines built environment stocks, we affirm the effect of built environment stock variable on CO 2 emission by proving that considering built environment stock variable can eliminate the asymmetric effect of GDP per capita growth and decline on CO 2 emission. These results quantitatively underline the role of built environment stocks in human development, future emission pathways, and relevant climate policy.