We analyze vertical price transmission in the German biodiesel market studying the relationship between rapeseed oil, soya oil and biodiesel prices. We focus on the period from summer 2002 to late 2007 during which the German biodiesel market developed into the largest market worldwide, mainly driven by political intervention. Tests on the adequacy of a traditional linear vector error correction model provide strong evidence against it and favor a regime-dependent model. We consider the Markov-switching vector error correction model which allows for parameter switching between regimes to be suitable for the question to be analyzed. We find two distinct regimes with differing error-correction behavior. Estimation results indicate that only rapeseed oil prices adjust deviations from the long-run equilibrium between the three prices. Error-correction is found to be strong until 2005; however, it substantially weakens in 2006 and 2007 due to market distorting policy measures such as blending obligations and norms requiring the use of rapeseed oil.