Research and development (R&D) collaborations between an innovator and her partner are often undertaken when neither party can bring the product to market individually, which precludes value creation without a joint effort. Yet, the uncertain nature of R&D complicates the monitoring of effort, and the resulting moral hazard reduces a collaboration’s value. Either party can avoid this outcome by acquiring the capability that is missing and then taking sole ownership of the project. That approach involves two types of risks: one related to whether the other party’s capability will be acquired and one related to how well it will be implemented (if acquired). We find that the extent of these two risks determines the optimality of delaying contracting or of signing contracts with buyout and buyback options, a baseball arbitration clause, or a novel reciprocal option. Baseball arbitration and reciprocal option clauses are unique in two ways. First, unlike typical options with predetermined strike prices, they allow either party to determine the buyout price at the time of their offer. Second, they allow the offer’s recipient to “turn the tables” on the other party. Although baseball arbitration and reciprocal option contracts both address inefficient joint development and product allocation, they exhibit their own inefficiencies that stem from the two parties’ strategic behavior. The best choice of contract is determined by trade-offs between these inefficiencies. Our model explores the similarities between the baseball arbitration and reciprocal option clauses, and we propose a modification to the reciprocal option contract that would increase its profitability. This paper was accepted by Terry Taylor, operations management.