Public-private partnerships offer potentially important opportunities for pro-poor agricultural research in developing countries. Yet in the international agricultural research community--and with regard to the Consultative Group on International Agricultural Research (CGIAR) itself--we see few examples of successful public-private partnerships, and fewer examples where such collaborations have contributed to food security, poverty reduction and economic growth. This study assesses the opportunities for, and challenges to, creating and sustaining public-private partnerships between the international agricultural research centers of the CGIAR and leading multinational, research-based agribusiness companies. The study hypothesizes that the willingness and ability of public agencies and private firms to enter into partnerships are constrained by fundamentally different incentive structures; by insufficient minimization of the costs and risks of collaboration; by an inability to overcome mutually negative perceptions; by limited use of creative organizational mechanisms that reduce competition over key assets and resources; and by insufficient access to information on successful partnership models. The study methodology is based on interviews and discussions with key stakeholders and a wide review of the literature on public-private partnership. Tentative findings suggest that while incentives and perceptions do differ between sectors, sufficient common space exists or can be created through incentive structuring to facilitate greater partnership. However, both public- and private-sector partners inadequately account for and minimize the costs and risks of partnership. Similarly, partners discount the need for brokers and third-party actors to manage research collaborations and reduce competition between sectors. Finally, partners are operating without sufficient information on existing partnership experiences, lessons, and models, potentially contributing to a persistent or widening gap between sectors.