The FAIR principles

FAIR data are useful data. Useful means that the data are findable, accessible, interoperable and reusable.

Making research data FAIR is a foundation in making science more open, transparent and profitable.

The FAIR principles fully respect that not all data can be entirely open – due to confidentiality or sensitivity, for example. Thus, ‘FAIRness’ can be applied in various degrees – from completely open and fair to closed but fair enough in a given context.


  • FAIR data can be discovered by both humans and machines, for instance by exposing meaningful metadata and search keywords to search engines and research data catalogues.
  • FAIR data can be found via unique and persistent identifiers, preferably DOIs. In this way, FAIR data can also be cited – just like a paper in a scientific journal.

Many researchers have traditionally presented their data and results on their own homepages. However, these often only appeal to a small group of colleagues. And they have the disadvantage of becoming outdated quickly. Furthermore, URLs are unstable and thus not suitable for referencing.

The easiest way of making data easily findable is publishing them in research data repositories that provide persistent identifiers and standard metadata including licenses for use of the data. Exposing data in a prominent place increases the chance that others become interested in that particular research – meaning higher impact, more citations and new collaborations.


  • FAIR data are preserved on a long-term storage, where they can be made available using “standard technical procedures”.
  • This does not mean that all data have to be openly available for everyone. But information on how the data can be retrieved have to be available, such as by signing a ‘terms of use’ statement or registration with a valid email address.
  • One can also consider publishing only the metadata and grant individual access to the full data upon personal request.

A standard technical procedure for giving access to data could for example be the provision of a collection (e.g. zip archive or BagIt container) for download including the original data, documentation and technical specifications of the data.

FAIR data are interoperable in three ways:

  • Technical interoperability means that files can be exchanged and used across different applications and systems – also in the future. This can be achieved by using widely accepted, sustainable and open standards and file formats. The metadata should contain information on which software had been used to create the data. If possible, special software that is needed to access and interpret the data should be archived along with the data.
  • Semantic interoperability means that the data can be understood and used in new contexts and research areas different from the one that it was created in originally. The data should also enable combination with other data sets, automatized analysis and machine learning. This is best done by using standard metadata schemes and providing all metadata needed to comprehend and interpret the data.
  • Legal interoperability means that there are clear terms and conditions on how the data may be accessed and reused, preferably by applying machine-readable standard licenses. This also prevents potential misuse of the data and makes sure that the original creator receives the right credit.

FAIR data helps to

  • build on previous research results
  • improve the quality of research by allowing others to assess the results
  • encourage collaboration and avoid duplication of effort
  • speed up innovation by exploiting synergies
  • involve citizens and society
  • improve transparency of the scientific process

Making data reusable means adding value to the data.

  • Provide all metadata and documentation needed for others to be able to reuse the data.
  • Provide clear terms of reuse through standard licenses.
  • Provide technical means for accessing and reusing the data.
  • When reusing data, respect the terms of use, acknowledge the creators and cite the original source properly.

More about FAIR