Abstract The goal of this paper is to develop a test for the relative importance of the time-varying term premium and the peso-problem for rejection of the Expectation Hypothesis of the Term Structure (EHTS). Our reasoning is based on a term structure model that allows for both phenomena simultaneously. If we assume that only one regime is observed ex post , we can estimate all the information we need to evaluate distortions generated by both hypotheses. We can also test the presence of a peso-problem. Firstly we find that a peso-problem might explain rejection of the EHTS in Germany and the United Kingdom after the European exchange rate crisis. Secondly, we show that this explanation appears inappropriate to explain the EHTS failure in the United States.