We analyze the optimal division of profit with complementary innovations. Even if each innovation can be achieved independently of the others, we identify circumstances in which the research should be conducted sequentially, targeting one innovation after another in a prespecified order. We then consider the implementation of this solution in a market equilibrium with specialized research firms. Firms are involved in a war of attrition, where each has an incentive to wait for the others to successfully complete their R&D projects before investing. To speed up innovation, the optimal policy must reward early innovators more generously than late ones.