Digital Object Identifier

A Digital Object Identifier, commonly abbreviated DOI, is a permanent alphanumeric string assigned to a digital document, dataset, or other resource so that it can be reliably cited and located regardless of changes to its underlying web address.

A DOI typically appears as a prefix identifying the registering organization followed by a suffix chosen by that organization, such as 10.1000/example. When a reader clicks a DOI link or enters it into a resolver, the request is redirected to the resource current web location. Because the publisher or depositor can update that redirect whenever the underlying address changes, the DOI itself never needs to change, which solves the long-standing problem of broken links in scholarly citation.

How DOI Registration Works

DOIs are issued through registration agencies operating under the International DOI Foundation. Crossref is the largest agency for scholarly journal articles and conference papers, while DataCite focuses on datasets, software, and other research outputs. Publishers and repositories pay a registration fee and submit metadata, such as title, author, and publication date, when they register a DOI, and that metadata becomes searchable through the registration agency services.

Use in Citation

DOIs have become a standard part of scholarly citation, appearing alongside or in place of traditional page-number citations in reference lists. Because the DOI resolves directly to the content, readers can follow a single stable link rather than hunting for a journal volume and page range, which is especially useful for born-digital content such as datasets or preprints that lack a conventional print analog.

Limitations

A DOI guarantees that a link resolves to some version of the resource, but it does not guarantee the content itself remains unchanged over time or that the hosting organization will exist indefinitely; long-term preservation still depends on the registering institution maintaining its infrastructure and honoring its commitment to keep the identifier resolvable.

Metadata and Interoperability

Beyond simply resolving to a web address, DOI metadata registered with agencies such as Crossref or DataCite powers a wide range of scholarly infrastructure, including citation tracking services that count how often a work has been referenced, and systems that link datasets to the articles that describe them. This metadata layer makes DOIs useful for far more than link resolution alone, functioning as a shared data backbone across publishing and research platforms.

Economic Considerations

Registering DOIs involves fees paid by publishers, repositories, or data centers to the relevant registration agency, with costs typically scaled to the volume of identifiers registered. These costs are usually absorbed into subscription or publication fees rather than charged directly to individual authors, though the cumulative expense across a large publishing operation can be substantial, particularly for data repositories issuing DOIs for many small datasets.

International Perspectives

DOI adoption varies by region and discipline, with registration agencies operating differently across countries and research sectors. Some national research infrastructures have established their own DOI-issuing bodies to serve local repositories and reduce dependence on international registration agencies, while still complying with the shared global technical standard that ensures all DOIs remain interoperable regardless of which agency issued them.

Practical Guidance

Researchers citing a work with a DOI should include the identifier in its standard resolvable form, typically as a full web address beginning with the DOI resolver prefix, rather than just the bare alphanumeric string, since this ensures the citation functions as a clickable link in digital formats and remains unambiguous in print.

Related Specialties

Working with DOIs and the broader scholarly identifier ecosystem draws on skills in metadata management, publishing technology, and scholarly communication, with specialized roles at publishers and repositories dedicated to registering identifiers correctly and ensuring metadata remains accurate as works are updated, corrected, or retracted after initial publication.

Looking Ahead

As research increasingly produces outputs beyond traditional articles, including software, datasets, and preregistered study protocols, DOI registration practices continue to expand to cover these newer output types, with registration agencies developing specialized metadata schemas suited to describing non-article research products accurately.

Handling Corrections and Retractions

Because a DOI is meant to remain permanently associated with a specific published work, publishers follow established conventions for handling corrections, versions, and retractions without breaking the existing identifier. A correction notice or retraction is typically linked to the original DOI through updated metadata rather than by issuing an entirely new identifier, preserving the citation trail while still alerting readers that the content status has changed since original publication.

This convention matters for scholarly integrity, since a reader following an older citation should be able to discover, through the DOI metadata, whether the work they are citing has since been corrected or withdrawn, rather than unknowingly relying on information the scholarly community has already flagged as unreliable.

Content Negotiation

Advanced DOI resolution supports content negotiation, allowing the same identifier to return different representations of metadata, such as machine-readable formats for software versus human-readable landing pages for browsers, depending on what the requesting system asks for, which makes DOIs useful infrastructure for both human researchers and automated systems alike.

Scale of Adoption

Hundreds of millions of DOIs have been registered across scholarly publishing and research data since the system origins in the late 1990s, reflecting its position as foundational infrastructure that underlies citation tracking, discovery services, and reference management software used throughout the research community.

This massive scale of adoption is precisely what makes the system reliable, since any widely used infrastructure of this kind benefits from the collective investment of thousands of participating organizations maintaining it.

Reference management software and citation tools now rely on this infrastructure as a basic assumption, automatically fetching metadata whenever a user enters a valid identifier into their library.

Editorially reviewed by Specialty Digest Editorial TeamLast reviewed November 3, 2026Researched and drafted with AI assistanceReport an issue