We examine how the efficiency of the judicial system impacts corporate innovation. To do so, we exploit a pilot program introduced by the U.S. Congress in 2011, which allowed judges with expertise(as opposed to randomly selected judges) to preside over more patent cases to facilitate efficient ruling. We find firms headquartered in counties subject to the Patent Pilot Program increase patent-based innovation by 5.2% to 6.2%, relative to firms in counties not under the program. Our results are concentrated among firms with high legal costs and uncertainty: firms that engage in innovation with “fuzzy boundaries”, that have high litigation risk, and that are more resource-constrained. However, we also find non-random assignment has an adverse impact on firms more likely to be assigned to judges that are favorably-biased towards non-practicing entities (NPEs) or “patent trolls”, who engage in frequent, frivolous litigation. Taken together, our findings underscore the important role of judicial efficiency in helping firms better allocate their resources towards innovation investment, but also indicate that judicial efficiency programs can exacerbate the negative effects of judicial biases in certain contexts.