The Crossref Curriculum

Descriptive metadata

Descriptive (bibliographic) metadata consists of metadata used to describe and cite an item. Examples of bibliographic metadata include authors, titles, pages, dates. The bibliographic metadata you register with Crossref is used in two significant ways:

  1. Matching DOIs to citations
  2. Capturing citations in reference management tools

Reference (or citation) management tools help researchers record, manage, and use bibliographic citations (references). The user adds references to the software, which uses the metadata of each source to create a database. Information from the database can be shared, and used in generating bibliographies. Learn more about different types of reference management software.

Our best practices are designed to maximize matching likelihood and ensure that bibliographic management tools are able to use our metadata to generate citations.

Learn more about the three main types of metadata: descriptive (bibliographic), administrative, and structural.

On this page, learn more about:


A contributor is a person or organization that is considered the author of a work. A contributor may be an author (default), editor, chair, or translator. Emphasize to your authors why you need to collect this information, explain how having accurate metadata will increase chances of accurate citation matching, credit the appropriate people fairly, and improve discoverability of their work. Learn more about collecting authenticated ORCID iDs in ORCID Collect & Connect for publishing organizations.

Including at least one author increases the chance a citation will be matched with a Crossref metadata record or DOI, and allows authors to receive credit for their work. Because our metadata is used for display and author attribution, order is important: flag first authors with the first attribute. Including ORCID iDs allows accurate identification of contributors, and means we can associate metadata records with contributions in ORCID.

  • Include at least one contributor, and you should include all contributors
  • List contributors in the appropriate order
  • Include complete given name(s), not just initials
  • If a contributor has just one name, put it under the <surname> field
  • Include ORCID iDs for all contributors when possible
  • Include affiliations for all contributors
  • For corporate authors, use the organization element
  • Don’t include suffixes such as Jr, Sr, IV in the family name field - use the suffix element


An author’s ORCID iD should be included whenever possible. Providing an ORCID iD in a metadata record allows the author’s ORCID record to be automatically updated via our auto-update process. OJS users who have upgraded to version 3.1.2 or later can request authenticated iDs from both contributing authors and co-authors - learn more about the OJS-ORCID plugin.

Include ORCID iDs in the element.

Note that the data supplied in the <given_name> and <surname> fields is used for display and query matching and must be as accurate as possible.

<person_name sequence="first" contributor_role="author">
  <ORCID authenticated="true">https\://</ORCID>

Contributor order

The <person_name> and <organization> elements both include required contributor role and sequence attributes. An author may be first or additional. Specific sequence numbering is not allowed, but many systems using Crossref metadata assume that the order of authors as present in the metadata is the appropriate order for metadata display.

Contributor example

  <person_name sequence="first" contributor_role="author">
    <affiliation>University of Cathair</affiliation>
    <ORCID authenticated="true">http\://</ORCID>
  <person_name sequence="additional" contributor_role="author">
    <given_name>Christopher </given_name>
    <affiliation>University of Cathair</affiliation>
  <person_name sequence="additional" contributor_role="author">
    <given_name>Katharine </given_name>


It is best to generate the metadata from your content management system and not from your typeset document, in order to avoid metadata reuse problems later, such as confused matching, messed up citation counts, and inconsistent display. Review how the title is treated or changed throughout the various stages of your production workflow. Follow these best practices to make sure your metadata can be used correctly by reference management tools.

  • Title must be in title or sentence case (not ALL CAPS)
  • Title field must not include other metadata such as author, price, volume numbers
  • Use separate title elements for different language titles - do not cram multiple titles in multiple languages into one element
  • Subtitles should be recorded in a separate subtitle element
  • Use UTF-8 encoding
  • Journal titles should reflect the journal title at the time of publication, and should not be updated if the journal title changes later on
  • May include face markup, LaTeX, or MathML where appropriate

If you need to update or correct a title, learn more about updating title records.


Publication dates are essential citation metadata: they must be included to match a citation to a DOI and vice versa. This also affects billing, as a back year must have a back year date at item level. Articles are cited at the item level, so the date in your metadata must be consistent with the date given in a citation for a match to be made.

  • Include a publication_date for every item registered
  • Include the correct date at both the parent and child level (journal issue / article, book title / chapter)
  • Include day, month, and year for each date (unless only month and year recorded)
  • Include both online and print publication dates (if applicable)


An International Standard Serial Number (ISSN) or International Standard Book Number (ISBN) is a number used to uniquely identify a serial or book publication. To obtain an ISSN, you need to register with the ISSN International Centre; and for an ISBN, with your national ISBN agency.

ISSNs/ISBNs are useful in distinguishing between serials or books with the same title. If a publication with the same content is published in more than one format, a different identifier is assigned to each media type. For example, a journal may have a print ISSN and an electronic ISSN, and print and ebooks have different ISBNs.

  • Include the title and ISSN/ISBN when you first deposit metadata for a content item in our system (if applicable)
  • Include both print and electronic ISSNs/ISBNs (if applicable)

If the journal does not have an ISSN at the time of registering content for it, include a title-level DOI for the journal. Once the ISSN is known, deposits should include both the ISSN and the journal-level DOI. Ideally, you would also update the metadata for all the previously registered content to include the ISSN. If you have any queries, please contact us.

We do not verify your title and ISSN combination with an external agency, but we carry out a check digit validation on every ISSN deposited. Once a title or ISSN is deposited, a new publication with the same title or ISSN can’t be created. If you try to make another deposit using a title and ISSN combination that does not match the combination in our system, the deposit will not work. Learn more about updating title records, including ISSNs/ISBNs.

Page or article numbers

  • Include first and last page numbers (if applicable)
  • Include article numbers (if applicable)

Journal title management

Titles in the Crossref system are created from the metadata you submit when you first register your content. You determine the exact title and ISSN included in the deposit, and we add that title and ISSN to our database. The title, ISSN, and title-level persistent identifier supplied in your content registration files must be consistent across submissions.

You are strongly encouraged to create a title-level persistent identifier (DOI) for your journal. An ISSN is valuable as a unique identifier of key title information – therefore our title management policies comply with ISSN center requirements for title management. Learn more about depositing ISSNs.

Practices that may affect registering your journal content include:

  • All series (journal, book, conference proceedings) should have an ISSN. The ISSN is crucial for identifying a serial. If you are supplying us with data for older titles that predate ISSN assignment, you should request ISSNs from your ISSN agency as they can be assigned retroactively. This isn’t only for our convenience - libraries, database providers, and other organizations using your data will welcome (and often require) an ISSN for anything defined as a journal.
  • The exception: new journal titles are sometimes not assigned an ISSN until some content has been published. Journal titles may be registered with Crossref without ISSNs if a journal title-level persistent identifier is included in the deposit. The title-level identifier must be supplied with every deposit for the journal until an ISSN is assigned.
  • A distinct ISSN should be supplied for each distinct version of a title. If a title changes significantly the publisher should obtain new ISSNs (both print and online). This rule is established by the ISSN International Centre, not Crossref, but we support and enforce it. Minor title changes (such as changing ‘and’ to ‘&’) don’t require a new ISSN.
    • Original title: Journal of Crossref Metadata
    • Minor change (no new ISSN required): The Journal of Crossref Metadata
    • Major change (new ISSN required): Journal of Crossref Metadata Quality
  • Online versions of journals encompassing multiple historic print titles should each be assigned distinct print and online ISSNs. Some publishers consolidate all versions of a title under the most recent title. This isn’t recommended practice as it causes a lot of linking and citing confusion – you’ve essentially created two (or more) versions of a title. This is particularly confusing when volume and issue numbers overlap between title iterations.

For recommendations on displaying information about journals, see the ISSN Manual, and NISO’s recommended practice on Presentation & Identification of E-Journals.

Journal-level DOIs

You are encouraged to create title-level DOIs for all journals (learn more about DOIs at different levels). You may also deposit DOIs at the issue and volume level. There are no content registration fees for journal-title-level, journal-volume-level, or journal-issue-level DOIs. When creating a journal-title-level DOI, note the following:

  • A distinct DOI should be created for each version of a title registered with Crossref. Any changes requiring a new ISSN should result in a new title-level DOI.
  • A journal-title-level DOI should resolve to a landing page that displays the same title and ISSN (when available) recorded in the Crossref database.
  • Once assigned, a journal-title-level DOI must be maintained.
  • Responsibility for maintaining a journal-title-level DOI transfers to the new owner when title ownership is transferred.
  • As with all content registered with Crossref, the DOI being resolved should appear on the DOI landing page.

Depositing title-level DOIs

Journal-title-level, journal-volume-level, and issue-level DOIs may be included with journal article deposits, or may be deposited separately. Once deposited, a journal-, volume-, or issue-level DOI does not need to be included in subsequent deposits but, as with all DOIs, the URL should be maintained. The web deposit form supports deposit of title-level DOIs. Example title-only deposit files are available here: with ISSN | without ISSN.


Abstracts imported from NLM JATS-formatted XML may be included in Crossref deposits. A namespace prefix (jats:) must be used for the abstract and all child elements, and the namespace must be included in the schema declaration. MathML may be included in abstracts but must use a MathML-specific namespace prefix. Multiple abstracts may be included.

Abstracts schema declaration

<doi_batch version="4.3.6" xmlns="" xmlns:xsi=""xsi:schemaLocation="" xmlns:jats="" **xmlns:mml="**" >

Example of a JATS-formatted abstract

<jats:abstract><jats:p>Acute and chronic lung inflammation is associated with numerous important disease pathologies including asthma, chronic obstructive pulmonary disease and silicosis. Lung fibroblasts are a novel and important target of anti-inflammatory therapy, as they orchestrate, respond to, and amplify inflammatory cascades and are the key cell in the pathogenesis of lung fibrosis. Peroxisome proliferator-activated receptor gamma (PPAR**<mml:math><mml:mi>**γ**</mml:mi></mml:math>**) ligands are small molecules that induce anti-inflammatory responses in a variety of tissues. Here, we report for the first time that PPAR**<mml:math><mml:mi>**γ**</mml:mi></mml:math>** ligands have potent anti-inflammatory effects on human lung fibroblasts. 2-cyano-3, 12-dioxoolean-1, 9-dien-28-oic acid (CDDO) and 15-deoxy-**<mml:math><mml:msup><mml:mi>**Δ**</mml:mi><mml:mrow><mml:mn>**12**</mml:mn><mml:mo>**,**</mml:mo><mml:mn>**14**</mml:mn></mml:mrow></mml:msup></mml:math>**-prostaglandin J<jats:sub>2</jats:sub> (15d-PGJ<jats:sub>2</jats:sub>) inhibit production of the inflammatory mediators interleukin-6 (IL-6), monocyte chemoattractant protein-1 (MCP-1), COX-2, and prostaglandin (PG)E<jats:sub>2</jats:sub> in primary human lung fibroblasts stimulated with either IL-1**<mml:math><mml:mi>**β**</mml:mi></mml:math>** or silica. The anti-inflammatory properties of these molecules are not blocked by the PPAR**<mml:math><mml:mi>**γ**</mml:mi></mml:math>** antagonist GW9662 and thus are largely PPAR**<mml:math><mml:mi>**γ**</mml:mi></mml:math>** independent. However, they are dependent on the presence of an electrophilic carbon. CDDO and 15d-PGJ<jats:sub>2</jats:sub>, but not rosiglitazone, inhibited NF-**<mml:math><mml:mi>**κ**</mml:mi></mml:math>**B activity. These results demonstrate that CDDO and 15d-PGJ<jats:sub>2</jats:sub> are potent attenuators of proinflammatory responses in lung fibroblasts and suggest that these molecules should be explored as the basis for novel, targeted anti-inflammatory therapies in the lung and other organs.</jats:p></jats:abstract>
<publication_date media_type="print">


Registering references means submitting them as part of your Crossref metadata deposit. It is optional, but strongly encouraged. Registering references is a prerequisite for participating in our Cited-by service, which shows the works which cite your work. Note that registering references is not the same as reference linking - learn more about the differences.

References registered as part of your metadata make your content more discoverable (especially if your references are open), enable evaluation of research, and helps with citation counts.

Whenever you register content with us, make sure you include your references in the submission:

Including references in your XML

When depositing references as part of your content registration XML, mark up individual citations according to our deposit schema section. For example:

<citation key="ref1">
  <journal_title>Current Opinion in Oncology</journal_title>

Marking up each element creates a structured citation. This allows us to be very precise when identifying potential matches with registered DOIs.

If you know the DOIs of individual citations, include them:

<citation key="ref2">

All citation elements are optional, but please submit as much information as possible to help us match your citations to DOIs accurately.

References may also be included as an unstructured citation. This option is not as precise as including an already-matched DOI or marking up a citation into individual elements:

<citation key="ref=3">
<unstructured_citation>Clow GD, McKay CP, Simmons Jr. GM, and Wharton RA, Jr. 1988. Climatological observations and predicted sublimation rates at Lake Hoare, Antarctica. Journal of Climate 1:715-728.</unstructured_citation>

Making your references open

All metadata that members register with us is publicly available (open) through our various metadata search and API services. However, references have historically been treated differently from all other metadata.

Our reference distribution policy means that members joining since 1 January 2018 have their references set to open by default, and are included in our full list of members who have opened their references for distribution.

Any member can choose to set their references to open, limited, or closed at any time. The policy can be set differently per DOI prefix.

Reference distribution options

References setting Description
Open Everyone can access your references through our metadata channels, including REST API and UNIXSD query results
Limited Your references can only be accessed by members participating in our Cited-by and Metadata Plus services (this is the default for older accounts)
Closed Your references are only used for our Cited-by service and are not distributed through any of our public Crossref interfaces

If you joined us before January 2018 and want to make your references publicly available (this is recommended), please contact us and we’ll enable reference distribution for your prefix (or prefixes). To enable open references for individual items, you can include a reference_distribution_opt flag within your <journal_article> element, for example: <journal_article publication_type="full_text" reference_distribution_opts="any">

To check if you (or your service provider) are including references in your metadata deposits, you can:

To find out if your references are currently set to open, limited, or closed, you can:

If you need further help, please contact us.

Last Updated: 2020 April 8 by Amanda Bartell