Lawrence Berkeley National Laboratory (LBNL) Residential Diagnostics Database (ResDB) contains blower door measurements and other diagnostic test results of homes in United States. Of these, approximately 134,000 single-family detached homes have sufficient information for the analysis of air leakage in relation to a number of housing characteristics. The authors performed regression analysis to consider the correlation between normalised leakage and a number of explanatory variables: International Energy Conservation Code (IECC) climate zone, floor area, height, year built, foundation type, duct location, and other characteristics. The regression model explains 68% of the observed variability in normalised leakage. ResDB also contains the before and after retrofit air leakage measurements of approximately 23,000 homes that participated in weatherization assistant programmes (WAPs) or residential energy efficiency programmes. The two types of programmes achieve rather similar reductions in normalised leakage: 30% for WAPs and 20% for other energy programmes.