Patent boxes significantly reduce the tax rate applied to income earned from a patent. Existing work finds that those reductions increase the number of patents. That said, not all patents are equally novel. In particular, the patent box encourages the submission of patents of marginal novelty, a selection effect that would reduce the average success rates of patents. At the same time, the increased return to patenting encourages additional effort in application preparation and prosecution, increasing success rates. While this predicts an ambiguous effect, due to lower financing costs, the net impact should be smaller for frequent innovators. We use data from applications to the European Patent Office from 1978 to 2017 and find that the introduction of a patent box increases the average success rate of applications by 4.4 percentage points, with the estimated effect becoming negative for frequent innovators. We further find that this effect is greater when boxes apply only to new innovations and when local development is required to access tax reductions. This suggests that for the frequent innovators, who form the bulk of submissions, patent boxes may indeed be encouraging the submission of marginally-novel applications.