How to Cite Preprint Published Version: Which One and How to Swap
Which version to cite when a preprint is later published, how to find the journal version on arXiv, bioRxiv and Crossref, and a worked APA 7 swap using APA's own example.
Wondering how to cite preprint published version or preprint in APA 7? If a journal has published the paper and you can read it, cite that version. APA says "Ideally, use and cite the final, published version of a work." If you only used the preprint, cite the preprint. Either way, cite the one whose words you actually relied on.
That rule comes from APA Style's preprint references page. The hard part is finding out whether a published version exists, because the links between the two are not always there. This guide shows where to look and works through a real case.
How to cite preprint published version or preprint: a decision table
Cite the version you read, and prefer the published one when you can read it. APA's page puts it in two sentences: use the final published version where possible, but "if you used the preprint version of a work, cite that version".
| Your situation | Which version to cite |
|---|---|
| A published version exists and you have read it | The published version, in journal article format |
| A published version exists but you only read the preprint | Read the published version, check your point still holds, then cite it. If you cannot get it, cite the preprint you used |
| The published version changed the finding you rely on | Cite the published version and report what it now says. Keep the preprint only if your point is about the change |
| Your point depends on something only in the preprint | Cite the preprint, with the version you read |
| No published version can be found | Cite the preprint, and record the date you checked |
| You read an author manuscript in a repository such as PubMed Central | APA treats this as a preprint format; its example cites PubMed Central as the archive |
The last row surprises people. APA notes that preprint versions "may or may not be peer reviewed", or "may be the author's final, peer-reviewed manuscript as accepted for publication". Its second example is a manuscript held in PubMed Central, cited with the preprint pattern.
Whether a preprint is fine to cite at all is a separate question. Our guide on whether you can cite a preprint in a thesis covers that, with university examples.
How to find out whether a preprint was published
Check four places, in this order: the preprint page, the preprint's DOI record, a title search, and the journal itself. A missing link in one place does not prove the paper is unpublished.
On arXiv: the journal reference field
arXiv lets authors add a journal reference and DOI after publication, and it says a DOI "will be linked from the arXiv abstract page". Its journal reference page says authors can add this "via the journal reference facility", and a good reference gives the "journal name, volume number, year and page number(s)". Adding it creates no new version.
arXiv also takes automatic updates from publishers. It says it accepts "feeds of DOI and journal reference metadata from publishers and other trusted third-parties". So the field can be filled by the author or by a feed, and it can also be empty for a paper that is published.
On bioRxiv: the link to the published version
bioRxiv's FAQ says it "will usually automatically add a link to the published version". It also says that "On rare occasions a match is not made for technical reasons." All preprint versions stay on the site after publication, so the preprint you cited will still open.
In Crossref: the preprint relationship
Crossref's posted content guide says the preprint publisher "will update their metadata to associate the posted content with the DOI" of the accepted manuscript or version of record. Crossref looks for matches by title and first author and emails the preprint's publisher when it finds one. The preprint's landing page must then link to the published version with a clear label.
In the metadata, the link is a relationship. Crossref's relationships guide lists "isPreprintOf, hasPreprint" as the preprint pair. You can see a record's metadata by opening https://api.crossref.org/works/ followed by the DOI, and looking at the "relation" field.
By title search
Search the exact preprint title in a scholarly database. A title can gain capitals or small edits in the journal, as in the example below, so search distinctive words and check the authors. The worked example below shows why this step matters.
Worked example: APA's own preprint example, now a journal article
APA's preprint example is a paper by Hampton and colleagues posted on PsyArXiv. Here it is, copied from APA's page:
Hampton, S., Rabagliati, H., Sorace, A., & Fletcher-Watson, S. (2017). Autism and bilingualism: A qualitative interview study of parents' perspectives and experiences. PsyArXiv. https://doi.org/10.31234/osf.io/76xfs
We checked whether it was published. The Crossref record for the preprint DOI lists it as a preprint posted on 17 March 2017, and its relation field is empty. So Crossref did not point to any journal version.
A title search in OpenAlex found a journal article with the same title, apart from capitals, and the same four authors. The Crossref record for that article confirms it: Journal of Speech, Language, and Hearing Research, volume 60, issue 2, pages 435 to 446, February 2017. Its relation field is also empty.
Here is the published version written in APA's journal article format. The article title goes into sentence case and loses its italics, and the journal name and volume take them instead:
Hampton, S., Rabagliati, H., Sorace, A., & Fletcher-Watson, S. (2017). Autism and bilingualism: A qualitative interview study of parents' perspectives and experiences. Journal of Speech, Language, and Hearing Research, 60(2), 435–446. https://doi.org/10.1044/2016_jslhr-l-15-0348
The in-text citation does not change: (Hampton et al., 2017). It would change if the two versions had different years, so always re-check the year when you swap.
The lesson is plain. Neither Crossref record links to the other, yet a published version exists. If you had stopped at the preprint's metadata, you would have cited the preprint when APA prefers the journal article.
What changes in the reference when you swap
The swap changes the format, not just the DOI. Use this checklist against APA's two example pages.
- Title: italic in the preprint reference; plain in the journal reference, where the journal name is italic instead.
- Source: the archive name (PsyArXiv, arXiv, bioRxiv) becomes the journal name, volume, issue and pages. APA says to "Always include the issue number for a journal article."
- DOI: use the journal DOI. APA's DOIs and URLs page says "Include a DOI for all works that have a DOI" and "Do not change the capitalization or punctuation of the DOI".
- Year: check it. The journal year can differ from the year the preprint was posted.
- Authors and title: check them. Author lists and titles can change between versions.
- Your sentence: remove any "not yet peer reviewed" wording you added for the preprint, once you cite the journal version.
- Your claim: find the passage in the journal version that supports it. If the number or wording changed, update your text.
Sometimes the journal has not yet given the paper a volume or pages. In that case, APA's journal page says to "use the format for an advance online publication".
Keeping the preprint: citing the exact version
If you cite the preprint, name the version you read. arXiv's versions page says authors can cite each version with "the full identifier that includes the version number", in the form arXiv:YYMM.NNNNNvX. Each public version is "a permanent part of the scientific record".
On bioRxiv, revisions keep the same DOI. Its FAQ says you can cite a particular version by adding the version-specific URL, such as one ending in v2.
Here is a fill-in template for a preprint you keep, following APA's pattern. Replace every bracketed placeholder:
[Surname], [Initial]., & [Surname], [Initial]. ([Year]). [Preprint title in sentence case]. [Archive name]. [https://doi.org/DOI or version-specific URL]
Should you cite both versions
Cite the one you used. APA's page gives one rule for each case and no format for citing both together.
There is one case where both can appear. If your point is the change itself, for example that a reported effect shrank after peer review, cite each version as its own reference. Then say in the sentence which version says what. Ask your supervisor first if your department has its own rule.
What Thesisroom does with this
Thesisroom does not format or swap references for you. Citation Check looks each reference up in eight registries: Crossref, OpenAlex, Semantic Scholar, PubMed, Retraction Watch, DOAJ, CORE and Unpaywall. It shows which references it could verify, which are retracted, and which need checking by hand.
Once your references are verified, Reference Watch re-checks them every week and alerts you when one changes, such as a retraction. For more on reading the identifier at the end of each reference, see our guide on what a DOI is in a reference.
Frequently asked questions
Should I cite the preprint or the published version?
The published version, if you can read it. APA says to "Ideally, use and cite the final, published version of a work." If you only used the preprint, APA says to cite the preprint.
How do I know if an arXiv preprint was published?
Look for a journal reference and a DOI link on the abstract page. Authors or publisher feeds can add them, but the field can be empty for a published paper. Search the title in a database as well.
Does the DOI change when a preprint is published?
Yes. The journal version has its own DOI, and Crossref asks preprint publishers to link the preprint to it. In the worked example, the PsyArXiv DOI starts with 10.31234 and the journal DOI with 10.1044. Use the journal DOI when you cite the journal version.
What if the published version has a different title?
Cite the published title, in sentence case, in journal article format. Check the authors too, and make sure the passage you rely on is still in the published text.
Do I need to update my thesis if a preprint I cited gets published?
APA prefers the published version, so update it if you can before you submit. Read the published version first and check your claim still matches it.
Is an accepted manuscript in a repository a preprint in APA?
APA's preprint page says preprint versions may be the author's final, peer-reviewed manuscript as accepted for publication. Its PubMed Central example is cited with the preprint pattern.
Sources
- APA Style: Preprint article references
- APA Style: Journal article references
- APA Style: DOIs and URLs
- arXiv: Adding journal reference and DOI
- arXiv: Automated DOI and journal reference updates from publishers
- arXiv: Versions
- bioRxiv: Frequently asked questions
- Crossref: Posted content (includes preprints)
- Crossref: Relationships
- Crossref metadata: Hampton et al. preprint, 10.31234/osf.io/76xfs
- Crossref metadata: Hampton et al. journal article, 10.1044/2016_jslhr-l-15-0348
- OpenAlex: title search for the Hampton et al. paper
How this guide was made: Thesisroom drafted this guide with the help of AI writing tools, then checked every fact, figure and source by hand against the references linked above.