This paper presents a new testing method for the scapegoat model of exchange rates. A number of steps are implemented to determine whether macro-fundamentals are scapegoats for the evolution of exchange rates. Estimation is conducted using a Bayesian Gibbs sampling approach applied to eight countries (five developed and three emerging) versus the USA over the period 2002Q1–2014Q4. The macro-fundamentals that we consider are real GDP growth, the inflation rate, the long-run nominal interest rate, and the current account to GDP ratio. We calculate the posterior probabilities that these macro-fundamentals are scapegoats. For the inflation rate, these probabilities are considerably higher than the imposed prior probabilities of ½ in five out of eight countries (in particular, the Anglo-Saxon economies).