Household borrowing and spending rise with house prices, particularly for leveraged households, but household spending is not consumption. We propose an alternative borrow-to-invest motive by which house price gains affect household spending on residential investment: rational, leveraged households have an incentive to make additional residential investments when house prices rise. We test this motive by comparing responses in different categories of spending across more and less leveraged households. We find strong evidence of the borrow-to-invest motive in UK data. Credit constraints matter through reducing access to leveraged returns and so reducing lifetime resources, rather than through consumption smoothing.