The Orphan Drug Act (ODA) was designed to spur the development of drugs for rare diseases. In principle, its design also incentivizes pharmaceutical firms to develop drugs for "rare" subdivisions of more prevalent diseases. I find that in response to this incentive, firms develop drugs for ODA-qualifying subdivisions of non-rare diseases. The impact in these tailored drug markets represents half of the total R&D response to the ODA. I also find that 10-percent of the innovation in subdivided disease drugs induced by the ODA would have been conducted without the policy. While modest in size, this inefficiency suggests that agency problems should be considered when designing innovation policy.