In this paper a novel survey dataset allows us to use a direct measure for credit constraints as well as a direct measure for the innovative activity of a firm to identify the effects of credit constraints on the innovation behaviour of firms. Furthermore, the design of the survey questions and the panel structure of the dataset allow us to avoid problems commonly difficult to solve such as the existence of forward looking adjustments in a world of expectations or mutual causation, and moreover to analyse potential asymmetries in the effects of above average and below average credit conditions. As opposed to many other papers we find clear evidence for a negative effect of credit constraints on the innovative activity of firms. In addition, we find that below average financing conditions restrict innovative activity, while above average financing conditions do not foster it. To explain this novel result we extend the usual theory of innovation activity by rigidities with respect to a firm's individual innovation capacity, which leads to a differentiation between a long run and a short run equilibrium in innovative output.