Financial frictions affect the way in which different components of GDP respond to a monetary policy shock. We embed the financial accelerator of Bernanke, Gertler and Gilchrist (1999) into a medium-scale Dynamic General Equilibrium model and evaluate the relative importance of financial frictions in explaining monetary transmission. Specifically, we match the impulse responses generated by the model with empirical impulse response functions obtained from a vector autoregression on US time series data. This allows us to provide estimates for the structural parameters of our model and judge the relevance of different model features. In addition, we propose a set of simple and instructive specification tests that can be used to assess the relative fit of various restricted models. Although our point estimates suggest some role for financial accelerator effects, they are actually of minor importance for the descriptive success of the model.