We propose a general specification to estimate the strength of the flight home effect, provide a decomposition into lender and recipient effects and countryby country measures. Flight home occurs when the change in domestic credit extended by domestic banks cannot be accounted for by recipient or lender effects. Based on break-adjusted BIS data, we find evidence of flight home for almost all banking systems with the notable exception of Canada, the US and Japan. In periods of relative calm, reversals of the home bias are small. The result is cumulative renationalization of bank lending.