Using a panel logit framework, the paper provides an estimate of the likelihood of a house price correction in 18 OECD countries. The analysis shows that a simple measure of the degree of house price overvaluation contains a lot of information about subsequent price reversals. Corrections are typically triggered by a sharp tightening in the monetary policy interest rate relative to a baseline level in each country. Two different assessments of the current and future baseline estimates of monetary policy interest rates are provided: a simple Taylor rule and one extracted from a term structure model. A case study based on the Canadian housing market is presented.