What Happens If You Cite a Retracted Paper? (2026 Guide)

A paper being retracted doesn't make it disappear. The PDF stays online, it stays indexed in Google Scholar and half a dozen other databases, and every citation already pointing to it keeps pointing to it. Unless a reader actively checks, a retracted study looks exactly like any other source in a reference list — same title, same authors, same confident formatting — even though the publisher has formally withdrawn its support for the paper's conclusions.

This is a different problem from an AI tool inventing a citation that never existed at all (covered in our guide on why ChatGPT citations are often wrong). A retracted paper is real. It was peer-reviewed, published, and cited in good faith by everyone who used it — right up until something in it was found to be false, fabricated, or fatally flawed. That makes it, in some ways, a harder trap to notice.

This guide explains what a retraction actually means, why retracted papers keep accumulating citations for years afterward, what actually happens when one slips into your own work, and the exact tools that catch it before that happens.

What Does "Retracted" Actually Mean?

A retraction is a formal statement from a journal or publisher withdrawing a previously published paper from the scientific record. It is different from a correction, which fixes a specific, contained error while the paper's overall conclusions stand, and different from an expression of concern, which flags that an investigation is underway but hasn't yet concluded. Once a paper is retracted, the publisher's position is that readers should no longer rely on it.

Retraction doesn't always mean an author acted in bad faith. Reasons range from deliberate data fabrication, image manipulation, and plagiarism, to honest mistakes — a mislabeled dataset, a statistical error that invalidates the conclusion, or accidental duplicate publication. What every retraction shares is this: the version of record no longer represents work the scientific community stands behind, regardless of the reason behind it.

The Retraction Watch Database — now maintained by Crossref — has logged more than 63,000 retracted papers, and the count grows every working day. On average, papers are retracted a little over two and a half years after publication, though some sit in the literature for decades before anyone catches the problem.

Why Retracted Papers Keep Getting Cited

Retracting a paper doesn't erase it from the internet, from citation databases, or from every reference list that already points to it. Multiple large-scale citation studies have found that retracted papers continue to accumulate citations for years, sometimes decades, after the retraction notice is published. One analysis of citations to retracted genetics papers found that nearly a quarter of all citations to 460 retracted articles arrived after the retraction had already happened. A separate analysis of roughly 3,000 retracted papers found that more than 30% of their combined 74,000 citations accumulated a year or more after retraction.

The more troubling number is how few citing authors even seem to notice. Across independent studies spanning biomedicine, computer science, and other fields, only somewhere between 4% and 7% of papers that cite a retracted study actually mention that it was retracted. In one large biomedical dataset of more than 13,000 post-retraction citation contexts, barely 5% acknowledged the retraction at all — the retracted paper was simply cited as if nothing had happened.

Part of the reason is structural. Google Scholar, still one of the most-used tools for literature searches, does not reliably flag retracted papers. Copies hosted on institutional repositories, preprint servers, and file-sharing sites frequently carry no retraction notice at all, even when the publisher's own version is clearly labeled. One analysis of retracted computer science papers found that while roughly 82% of official published versions were clearly labeled as retracted, researchers relying on secondary copies or general search were far more likely to miss the warning entirely.

A Real Example of Why This Matters

The consequences of an unnoticed retraction aren't hypothetical. In February 1998, The Lancet published a paper by Andrew Wakefield and twelve co-authors claiming a link between the MMR vaccine and autism, based on just 12 children. An investigation by journalist Brian Deer later uncovered manipulated patient data and an undisclosed financial conflict — Wakefield had been paid by lawyers preparing litigation against vaccine manufacturers. The UK General Medical Council found Wakefield guilty of serious professional misconduct in January 2010, and The Lancet fully retracted the paper weeks later — a full twelve years after publication.

By then, the paper had already been credited with a measurable drop in UK vaccination rates, and it continues to fuel vaccine hesitancy internationally today, long after every one of its central claims was discredited. The retraction did not undo the citation trail; it simply marked the point where the scientific record caught up with what had already spread.

A quieter, more routine example makes the same point on a smaller scale: a clinical trial report on an omega-3 nutrition intervention, retracted in 2008 after an investigation found the underlying data had been falsified, was still being cited positively — with no mention of its retraction — eleven years later. It had simply been picked up by researchers who never checked whether the study behind their own citation still stood.

What Actually Happens If You Cite a Retracted Paper?

The consequences scale with the stakes of what you're writing and why the paper was retracted in the first place.

As a student

At minimum, an instructor who checks sources closely may flag it as a research-quality issue. If the retraction was for fraud or fabricated data and your argument leans on that paper's findings, the foundation of your own claim is compromised — not just the citation formatting.

In a submitted manuscript

Journals increasingly check reference lists against retraction databases during peer review. A citation to a retracted paper, especially one retracted for misconduct, can trigger a request for revision or, in serious cases, rejection.

In an already-published paper

If discovered after publication, the journal may require you to issue a correction. If the retracted study was central to your conclusions, your own paper's validity can be publicly questioned or formally investigated.

In a systematic review or meta-analysis

This is the highest-stakes case. Including a retracted study's data in a pooled analysis can skew the field's understanding of an entire question, and has itself been the direct cause of subsequent retractions of the review that included it.

In clinical or professional practice

Citing, or worse relying on, a retracted clinical study to support a treatment recommendation carries the most serious real-world stakes of any case on this list.

How to Check If a Paper Has Been Retracted

Retraction Watch Database

The most comprehensive source available. Search directly by title, author, journal, or DOI — most other retraction-checking tools ultimately draw from this same dataset.

CrossRef metadata (via DOI)

For any DOI, CrossRef's record can carry an "update-to" relation pointing to a retraction or correction notice — a quick way to check a single source directly at the registry level.

Zotero and EndNote

Both reference managers integrate Retraction Watch data directly, flagging retracted items already in your library and warning you if you try to cite one — even retroactively, if something you cited earlier is retracted later. Coverage is generally limited to items with a DOI or PMID.

PubMed

For medical and health literature, retraction notices are linked directly to the indexed article record, making PubMed one of the more reliable single sources to check for that domain.

What not to rely on alone

Google Scholar and general web search do not reliably surface retraction status. Treat a clean search result as inconclusive, not as confirmation, for any source that matters to your argument.

Step-by-Step: Screening Your Reference List

1
List every source with its DOI where available
A DOI makes automated checking dramatically faster and more reliable than a title-only search.
2
Let your reference manager check first
If you use Zotero or EndNote, confirm the Retraction Watch integration is active and review any flagged items.
3
Search unmanaged sources directly in the Retraction Watch Database
For anything not in a reference manager, search it by title or DOI individually.
4
Cross-check PubMed for medical and health sources
Retraction and correction notices are linked prominently there for biomedical literature.
5
Read the actual notice if you find one
A correction or partial retraction may leave the specific point you're citing untouched; a full retraction for fabricated data does not.
6
Re-check long-running projects close to submission
A thesis, systematic review, or book manuscript can take years to finish — long enough for an early source to be retracted before you're done.
7
Run a final automated pass before submission
Upload your finished reference list to a reference checker as a last safety net across the whole document at once.

What to Do If You Find a Retracted Citation

If the specific point you cited is unaffected

A retraction for an unrelated, contained issue may still leave the exact fact you cited intact. In that case, it can be acceptable to keep the citation — but disclose the retraction in your text and explain why the specific point still stands.

If it's central to your argument, or retracted for misconduct

Remove it and find a replacement source. Do not keep a citation whose underlying data was fabricated or falsified, even if removing it weakens your paragraph — the alternative is worse.

If you've already published

Don't simply delete the reference silently in a future draft. Contact your editor or publisher about issuing a correction if the retraction affects a claim in your already-published work.

Frequently Asked Questions

Does citing a retracted paper automatically mean my work is wrong?

Not always. It depends on why the paper was retracted and whether the specific point you cited depends on the retracted finding. A paper retracted for an unrelated administrative error may not affect a citation used only for general background, but a paper retracted for fabricated data undermines anything drawn from its results.

Can I still cite a retracted paper at all?

Yes, in some contexts. Citing a retracted paper to discuss the retraction itself, as a case study of research misconduct, or as part of a history-of-the-field narrative is legitimate and common. The requirement is disclosure: make clear in your text that the paper has been retracted and why.

How is a correction different from a retraction?

A correction fixes a specific, contained error while the paper's overall conclusions stand. A retraction means the publisher no longer stands behind the paper as part of the scientific record. An expression of concern sits in between: a signal that an investigation is underway but not yet resolved.

Will my reference manager warn me automatically about retractions?

Zotero and EndNote both check items against the Retraction Watch Database and can warn you at the point of citing, but this coverage is generally limited to items with a DOI or PMID and isn't a substitute for checking sources that live outside your reference manager.

What's the fastest way to check an entire reference list at once?

Upload the full document to a dedicated reference checker rather than searching each source individually. A reference checker can flag sources that no longer resolve cleanly across an entire bibliography in one pass.

Conclusion

A retraction is supposed to be the moment the scientific record corrects itself. In practice, it's often just the moment the correction becomes available — whether anyone reading the paper afterward actually finds it is a separate question entirely, and the data says most don't. That gap between "the correction exists" and "the correction reached the people still citing the paper" is exactly where an unnoticed retraction can quietly undermine a thesis, a manuscript, or a clinical recommendation years later.

Closing that gap doesn't require much: a DOI check against CrossRef, a search in the Retraction Watch Database, or a reference manager configured to warn you automatically. Done as you write rather than saved for the end, it takes minutes — and it's the difference between a reference list you can stand behind and one that quietly rests on work the scientific community has already withdrawn.

Screen Your References for Retractions Now

Upload your document and check every citation for retractions, fabricated sources, and formatting errors — before your instructor, editor, or a reviewer does.

Check My References

What Happens If You Cite a Retracted Paper? (And How to Avoid It)

Quick Answer: Citing a Retracted Paper