This paper presents an estimated DSGE model for the European Monetary Union. Our approach, contrary to the previous studies, accounts for heterogeneity within the euro area. We advance the empirical literature by estimating an open-economy model with unfiltered data, which is a much more challenging task than a similar exercise done in the closedeconomy framework. In the estimation we utilize disaggregated information, employing single country data, along with the aggregated EMU data by Fagan et. al (2001). We also contribute to the literature by proposing a strategy for consistent estimation of the currency union model, using information available prior to the adoption of the single currency and afterwards. This approach requires the determination of two separate data generating processes - here these are theoretical DSGE models - corresponding to both current and historical monetary regimes. We emphasize the use of regime-switching models in the DSGE framework (in our case the threshold is known exactly and the switch is permanent). The approach is illustrated by developing a simple tworegion DSGE model, with a particular focus on analyzing the German economy within EMU, and its Bayesian estimation on the sample 1980:q1- 2003:q4. Moreover, the paper offers: (i) a robustness check of the estimation results with respect to the alternative data approaches and various restrictions imposed on the model's structure, (ii) assessments of the relative importance of various shocks and frictions for explaining the model dynamics and (iii) an evaluation of the model's empirical properties.