This Article argues that courts have created a de facto extra-statutory condition of patentability, herein termed the" completeness" requirement. This requirement bars patents on certain inventions whose chief value lies in their function as inputs into downstream research. The Article contends that the notion of completeness explains doctrinal innovations that are difficult to rationalize any other way. Although it reflects an important policy of limiting unduly preemptive patent claims on foundational, building-block inventions, the completeness requirement in its current form fails to implement this policy in a way that is coherent and consistent with patent law's utilitarian goals. In addition, courts' attempts to develop the completeness requirement based on existing statutory provisions have resulted in controversial interpretations of the Patent Act, generating legitimacy costs. The Article argues that these problems are best addressed by explicitly recognizing completeness as a separate requirement of patentability and modifying the doctrinal tools used to enforce this requirement. To determine whether a patent claim passes completeness, the Article proposes a new test that focuses on the generality and unpredictability of a claimed invention's applications. Further, it argues that an amendment to the Patent Act codifying the requirement of completeness is the most effective way to implement the proposal. Finally, the Article explores the possibility of awarding a limited patent right, which it terms" Research Patent," to claims that satisfy existing requirements of patentability, but fail completeness. This right would provide the intellectual property incentives that are likely needed to develop and commercialize foundational inventions, and also help decrease the potential for stifling downstream innovation created by granting full patent protection to such inventions.