I study optimal incentive schemes in organizations where agents perform their tasks sequentially. I consider a model in which agents ' effort decisions are mapped into the probability of the project 's success. An optimal investment-inducing mechanism allocates rewards to agents so as to induce all of them to exert effort in equilibrium at minimal cost to the principal. I characterize the unique optimal mechanism in several versions of my benchmark model. I also address the problem of allocating individuals with diverse qualifications to different slots of the production process as well as allocating tasks of different importance across different agents. Copyright © 2006, RAND.