Retraction Checker
Paste a reference list in any style, or a column of DOIs, and every DOI in it is checked against Crossref's record of retractions, expressions of concern, corrections and withdrawals — including the Retraction Watch data Crossref opened to everyone in 2023. You get each work's status and every notice, with its date, its source and a link to read it. Free, no sign-up.
doi:10.…, DOI: 10.…, https://doi.org/10.… and bare 10.… are all found. Up to 100 DOIs per check.
Only the DOIs leave this page: your browser sends them to Crossref's public API. The rest of your list is not sent anywhere.
Paste a reference list or some DOIs, then check them.
How to Use the Retraction Checker
- 1. Paste your references. A whole list in any style, or DOIs one per line. A DOI is found however it is written —
doi:10…,DOI: 10…,https://doi.org/10…or bare — up to 100 per check. - 2. Press Check for retractions. Only the DOIs leave the page: your browser sends them to Crossref, twenty to a request.
- 3. Read the table. Each DOI gets a status and, under Notices, every notice Crossref holds for it: the type, the date, who recorded it, and a link to the notice. Tick Show only DOIs with a notice to hide the rest.
- 4. Act on what you find. Open each notice before you decide what to do with the source, and download the CSV if you need a record of the check.
Where the Notices Come From
A journal that retracts, corrects or raises concerns about an article usually publishes a notice with its own DOI and registers it with Crossref as an update to the original. Many retractions were never registered that way — the gap the Retraction Watch database fills. Crossref acquired that database in September 2023, opened it to everyone and merged it into its public API, where each work carries an updated-by list. The checker reads that list: each entry gives the notice's DOI, type and date, and its source — the publisher, or Retraction Watch with its record number in the Retraction Watch Database.
| Status | What Crossref records | What to do |
|---|---|---|
| Retracted | A retraction notice | Do not cite the work as evidence. |
| Withdrawn | A withdrawal or removal, often with the article page replaced by the notice | Treat it as retracted. |
| Partly retracted | A partial retraction | Read the notice before citing any part of the work. |
| Expression of concern | An editors' warning while a problem is investigated | Not a retraction, but do not lean on the work alone. |
| Corrected | A correction, erratum or corrigendum | Check that the correction does not touch what you cite. |
| Other update | A new version, addendum or clarification | Cite the version you read. |
| No notice found | Nothing | No notice in Crossref — not proof that the work is sound. |
| Is itself a notice | The DOI is a notice's, not an article's | Expected when you cite a retraction notice. |
When a work has several notices, the most serious decides its status: a paper that drew an expression of concern and was later retracted shows as Retracted, with both notices listed in date order.
A Retracted Paper, Checked
APA's own example of a retracted journal article is Joly, Stapel and Lindenberg (2008), one of the papers retracted after the Diederik Stapel data-fabrication investigation. Checked by its DOI, it comes back like this:
Result for 10.1177/0146167208318401
Title
Silence and Table Manners: When Environments Activate Norms — Personality and Social Psychology Bulletin, 2008
Status
Retracted
Notice
Retraction · Sep 18, 2012 · Retraction Watch, record 2855 · 10.1177/0146167212462821
Notice
Correction · Sep 18, 2012 · Publisher · 10.1177/0146167212462821
Both rows point to the same notice: Retraction Watch classes it as a retraction, the publisher's deposit as a correction. The checker keeps both, and the more serious sets the status, so a disagreement never hides a retraction. The notice itself, one click away, settles it.
What to Do With a Retracted Source
A retraction does not take a paper out of circulation: it stays in databases and in other authors' reference lists, often unmarked. Citing it as support repeats a result its own journal has withdrawn — an error a reviewer can find in seconds.
- Remove it when it was one support among several. Find a sound source for the claim, or drop the claim.
- Cite it as retracted when the retraction is part of your point — the history of a debate, a study of misconduct. Say so in the text, and give the retraction notice in your references too.
- Add a caveat for an expression of concern, and check back later: many end in a retraction or a correction.
APA 7: a retracted article, with the retraction noted
Joly, J. F., Stapel, D. A., & Lindenberg, S. M. (2008). Silence and table manners: When environments activate norms. Personality and Social Psychology Bulletin, 34(8), 1047–1056. https://doi.org/10.1177/0146167208318401 (Retraction published 2012, Personality and Social Psychology Bulletin, 38[10], 1378)
What “No notice found” can and cannot tell you
It means Crossref holds no notice for the DOI. Coverage depends on what publishers deposit and what Retraction Watch has recorded, and Crossref says its data on corrections and expressions of concern is less complete than on retractions. The checker lists notices but does not judge whether a later one reversed an earlier one, so a paper retracted and later reinstated still shows as retracted. For a source your argument depends on, read the publisher's page and search the Retraction Watch Database.
Frequently Asked Questions
- How do I check if a paper has been retracted?
- Look it up by DOI. Paste the DOI, or your whole reference list, into the checker above to see any retraction, expression of concern, correction or withdrawal recorded in Crossref, Retraction Watch's data included. Then open the publisher's page: a retracted article is usually watermarked or replaced by its notice, which says why.
- What is the Retraction Watch Database?
- A database of retractions built by the Center for Scientific Integrity, the nonprofit behind the Retraction Watch blog. Crossref acquired it in September 2023 and made it openly available: it is published as a CSV file updated every working day and merged into Crossref's API, which is where this checker reads it.
- What is an expression of concern?
- A notice from a journal's editors that they have doubts about a published work — its data, its methods or its ethics — while the matter is investigated. It is not a retraction, and it may later be followed by one, by a correction, or by nothing. If you cite the work, say that it carries an expression of concern.
- Can I cite a retracted paper?
- Only when the retraction is part of what you are writing about, such as a review that has to account for the paper. Say in your text that the work was retracted and cite the retraction notice as well. Never cite a retracted paper as support for its findings.
- How do I cite a retracted article in APA 7?
- Give the article's usual reference, then the retraction's details in parentheses after the DOI, as in APA's own example: "(Retraction published 2012, Personality and Social Psychology Bulletin, 38[10], 1378)". To cite the notice itself, reference it like any journal article, under the notice's own title.
- Why does a DOI show as Not in Crossref?
- Crossref is one of several DOI registration agencies. DOIs for most datasets and many preprints, arXiv's among them, are registered with DataCite, and Crossref holds no notices for them. Check the DOI for typos first; if it resolves at doi.org, search the Retraction Watch Database for the work's title instead.