Several empirical minimum wage studies have recently been published that simulate employment effects of a federal minimum wage in Germany. We disentangle various factors that explain the variation in previous simulation results. Based on data from the German Socio-Economic Panel and the newly available Verdienststrukturerhebung 2006 we conduct robustness analyses that systematically test the range in the outcomes of different labor demand simulations. We find that labor demand effects are sensitive to measurement errors in wages, the representativeness of the sample with respect to several types of labor inputs as well as estimated and assumed labor demand and output price elasticities. Interactions of those determinants may lead to substantial differences in simulation outcomes.