This paper discusses the advantages of Bayesian New Neoclassical Synthesis models as tools for monetary policy analysis and forecasting. The combination of a sound, micro founded structure with a good probabilistic description of the observed data makes those models suitable for investigating the structural sources of business cycle fluctuations, for analysing optimal monetary policy responses to those developments and for making economic projections conditional on various policy assumptions. The paper gives two examples of such analysis. (JEL: E40, E50, C11) Copyright (c) 2005 The European Economic Association.