The FAIR principles formulate guidelines for the sustainable reusability of research data. FAIR stands for Findability, Accessibility, Interoperability, and Reusability of data and metadata. While there is a growing body of general implementation guidelines, so far there is a lack of specific recommendations on how to apply the FAIR principles to the specific needs of social, behavioural and economic science data. These disciplines work with highly diverse data types that often contain confidential information on individuals, companies, or institutions. These features pose some challenges to the useful implementation of the FAIR principles – especially regarding the machine-actionability of data and metadata that is at the core of the FAIR principles. This White Paper defines the FAIR principles for the social, behavioural and economic sciences. For each of the 15 FAIR (sub)principles, the paper proposes minimum requirements and provides a vision for a full-implementation of the FAIR principles by repositories and data centres. The paper was authored by members of the Economic and Social Sciences goINg FAIR Implementation Network (EcoSoc-IN) and addresses research data centres and other stakeholders who strive for a FAIR research data infrastructure in the disciplines of KonsortSWD.