The world’s food systems must be transformed to make healthier diets more accessible for all, while increasing the sustainability and resilience of these systems, as clearly called for in the 2030 Agenda for Sustainable Development. Scientists and policymakers are among the complex set of actors necessary for this transformation. This brief reviews the importance of the interfaces between science and policy and how these interfaces can become more active and effective.

On the science side, there is a polarity, and sometimes tension, between “research-driven” and “demand-driven” research; competition among disciplines and approaches; strong private research, mainly business-focused, alongside public research focused largely on public goods; and an ongoing debate about legitimacy, excellence, and impact.

The world of policy is also diverse, including political actors shaping future visions and competing for governing roles; and public actors and policymakers, at many different territorial levels and sometimes focused on specific sectoral interests (such as education, health, and agriculture). Policy decisions are sometimes based on prevailing science, which serves as a key element of accountability and efficiency, but this is far from being the norm.

Research and policymaking interact in various ways. First, there is an overall relation of supervision between science and policy: research is governed and influenced by the state through continuous negotiation about orientation, budgets, and demand for creativity and freedom to explore new ideas. Second, numerous initiatives and organizations link governments and scientific institutions at national and local levels. At the international and multilateral level, there is a growing effort to build collective expertise to formulate state-of-the-art scientific knowledge regarding specific global problems. The aim is to identify and build consensus around legitimate and efficient political actions, to be agreed at global level and implemented at all levels.

On the whole, science has been very influential in forming consensus views on many topics linked to food systems and, from there, the orientation of policies. However, the gap between the scientific process for producing knowledge on a specific question and the complex process of policymaking, which must balance empirical information and scientific results with management of trade-offs, political agendas, and societal acceptability, points to the limitations to evidence-informed policymaking. Furthermore, policymakers and scientists are not the only players in these interfaces; many other stakeholders play an explicit or implicit, visible or invisible role, and power imbalances among them may be strong.

To improve the functioning of these science–policy interfaces, there is a call for both science and policy actors to go beyond their conventional roles. For scientists, the recommendation is to move beyond knowledge supply and alarm-bell ringing to become real knowledge brokers, to engage with policymakers and with key food systems actors, and to promote coalitions for change to co-design the future. For policymakers, the recommendation is to make more effective use of knowledge for decision-making by inviting scientists to deliberative dialogue processes, increasing understanding of uncertainty, complexity, and the limits of evidence, and making their expectations more explicit to the science community. This will require capacity-building for both sides.

There is not one science–policy interface but rather many, at different scales, for different functions, addressing different challenges. Strengthening, connecting, and streamlining these interfaces can ensure the consistency and success of food system transformation. To improve science–policy interfaces, the scientific community should:

(1) generate actionable knowledge, data, and metrics to move beyond obstacles, and address trade-offs and barriers to change, including power asymmetry, path dependency, conflicts of interest, and risk and uncertainty;

(2) articulate models, knowledge, and place-based innovations to design, implement, and assess specific transformative pathways—this requires specific arrangements, dialogues, and approaches, including scientific ones;

(3) connect expertise mechanisms, such as scientific committees, to address multisectoral and multiscale processes for sustainable development; and

(4) strengthen scientific cooperation through major challenge-oriented alliances and programs.

Science–policy interfaces can play a decisive role if they are able to dovetail divergent views and overcome polarized debates and sectoral fragmentation. These interfaces must also help us to look ahead and to bridge local and global processes and actions.