How to Find a DOI and Check That It Works

# How to Find a DOI and Check That It Works

A DOI (Digital Object Identifier) is a permanent identifier for a publication, written as a prefix that starts with "10.", a slash, and a suffix, such as 10.1126/science.1258096. If you have a DOI, put doi.org/ in front of it to open the article. If you only have a title and authors, search Crossref to find the DOI, then open it at doi.org to confirm it points to the same paper (DOI Foundation, n.d.; Crossref, n.d.).

First decide which case you are in: you have a DOI, or you don't. The steps differ.

What a DOI is and how it is built

The DOI Foundation defines a DOI name as "a digital identifier of an object, any object — physical, digital, or abstract" (DOI Foundation, n.d.). For a reference list, that object is usually an article, book, or dataset.

Publishers register DOIs when they publish. For journal articles that mostly happens through Crossref. Authors don't apply for a DOI themselves.

If you have a DOI: resolve it at doi.org

The DOI Foundation explains that putting doi.org/ before a DOI turns it into a link, for example https://doi.org/10.5334/dsj-2017-040. Clicking it "resolves" the DOI, which means it redirects to the latest information about the item, even if the item changes or moves (DOI Foundation, n.d.).

You can do this in two ways:

1. Go to doi.org, paste the full DOI (prefix and suffix) into the resolver box, and submit.
2. Type https://doi.org/ followed by the DOI into your browser's address bar. For example: https://doi.org/10.1126/science.1258096.

You land on the publisher's official article page. That doesn't mean the full text is free. Whether you can download it depends on whether the journal is open access and whether your institution subscribes.

If you only have a title and authors: find the DOI

There are three routes:

1. Crossref metadata search. Open search.crossref.org and enter the title and authors. Results cover journal articles, books, standards, and datasets, and each result shows its DOI. This is the most direct way to get a DOI (Crossref, n.d.).
2. PubMed, for medicine and life sciences. The record page shows the DOI. The PMID on the same page is PubMed's own number, not a DOI.
3. Google Scholar and the journal's website. Scholar's citation pop-up, and the top of a journal article page or its footnotes, usually list the DOI.

Once you have a DOI, resolve it at doi.org one more time. Check that the title, authors, year, and volume/issue match your reference before you use it.

When a DOI doesn't work

A DOI that won't resolve is a signal to investigate, not something to skip. Work through it in this order:

1. Look for a typo. Check the slash.
2. Then ask whether the reference exists at all. AI-generated citations often contain made-up DOIs, so a DOI that fails to resolve may mean the reference itself may not exist.
3. Fix or remove it. Either add the correct source or delete the entry.

If the DOI resolves but you can't download the paper, the DOI is probably fine. The limit is access, not the identifier.

Check a whole reference list at once

Doing this one DOI at a time gets slow with a long list. PerfectRefs works on the whole list: paste the reference list, with no need to upload the paper.

• Existence check looks each reference up in Crossref, OpenAlex, and other scholarly databases. It comes with a free daily allowance.
• Style check compares authors, title, journal, year, volume/issue/pages, and the DOI field with the database record, field by field. It shows the record's version with a source link, and you can accept a fix with one click. It uses premium credits (2 per reference), and new accounts get 10 premium credits.

Common mistakes

1. Treating a DOI as a URL. A web address can break. A DOI resolves to the item's latest location, so it still works if the article moves (DOI Foundation, n.d.).
2. Assuming doi.org means free download. It leads to the publisher's page, and access depends on the journal and your institution.
3. Skipping a DOI that won't resolve. Check for a copying error first, then check whether the reference exists.
4. Confusing a DOI with a PMID. They are different numbers (see the FAQ).

FAQ

What is a DOI? A DOI is a permanent digital identifier for a publication. It starts with 10., then a prefix, a slash, and a suffix, as in 10.1126/science.1258096. Publishers register it at publication, and it resolves to the article's official page even if the web address changes.

Where can I look up a DOI? If you have the DOI, resolve it at doi.org. If you only have the title and authors, search Crossref at search.crossref.org. PubMed, Google Scholar, and the journal's website also list DOIs.

How do I find a DOI from a title? Search the title and authors in Crossref, copy the DOI from the matching result, then open it at doi.org. Use it only if the title, authors, year, and volume/issue match your reference.

What is the difference between a DOI and a PMID? A PMID is the unique number PubMed gives each record. A DOI is a global identifier for a publication. NCBI notes that the unique identifier in the PubMed database is called the PMID (National Center for Biotechnology Information, n.d.). A PubMed record page shows both.

Does a DOI let me download the paper? Not necessarily. A DOI only takes you to the article page. Whether the full text is free depends on whether the journal is open access, whether your institution subscribes, or whether the paper is available as a public preprint.