About half, probably more, of all patented inventions in the United States are never commercially exploited. Even many of the most commercially significant inventions take decades to come to market. In this paper, I contend that the patent system is substantially retarding the commercialization of valuable inventions. This neglect should not come as a surprise - the dominant framework under-girding patent law, the "reward" theory, is premised on providing incentives for nascent inventions, not commercialized end-products. Although more recent "prospect" theories properly recognize the importance of patent protection for commercializing inventions, they incorrectly conclude that strong, real property-like rights for inventors are necessary to spur robust commercialization — sometimes, weaker rights are preferable. In analyzing these dominant theories of patent law, I conclude that it is effectively impossible to adjust the timing, duration, and scope of traditional patent rights in order to encourage substantial commercialization. In place of efforts to reform the traditional patent, whose quid pro quo is the disclosure of new and non-obvious information, I propose a new "commercialization" patent, granted in exchange for the commitment ro make and sell a substantially novel product. Decoupling the invention and commercialization functions of patent law into dual rights would yield more commercialization than the existing system, without unduly decreasing competition, encouraging rent-seeking, or increasing administrative costs.