We identify measures of shocks to total factor productivity and preferences from two real business cycle models and subject them to Granger causality tests to see whether they can be considered exogenous to other plausible sources of the German business cycle. For the period 60.i to 89.iv no variable Granger causes the shock measures, and for the period 70.i to 01.iv, only M3 does. We attribute the latter result to the breaks in our time series associated with the German reunification in 1990 and the European Monetary Union in 1999. We, thus, find no evidence to reject the exogeneity of our shock measures. Our findings contrast with similar studies for other countries that question the exogeneity of either productivity or preference shocks.