How to Cite a Retracted Article Transparently
Before citing a retracted article, read the retraction notice and decide why the article belongs in your paper. Do not present withdrawn findings as dependable evidence. If you discuss the retraction itself, identify that status clearly, distinguish the article from its notice, and follow your assigned citation style or journal rules.
plagiarism-checker-online.net Editorial Team | Sources checked October 8, 2026
You reopen a source before submitting your research paper and find a retraction banner. Your reference may still have the right authors, year, and DOI. The problem is the sentence it supports. Adding a label to the bibliography won't fix an argument that still treats the withdrawn result as established evidence.
This guide explains how to cite a retracted article while checking what needs to change in your paper. For ordinary source-credit decisions, start with the citation and referencing guide.
First decide what role the paper plays
Supporting a factual claim versus analyzing a retraction
Write the reason you're using the article in one sentence. “This paper demonstrates that the intervention works” is a different use from “This case shows how a journal explained a retraction.” The first relies on the findings. The second examines the publication record.
The ICMJE recommendations for medical-journal submissions tell authors to check references for retractions and avoid citing retracted articles except when referring to the retraction. That is medical-journal guidance, not a rule that automatically governs every college assignment. It gives you a useful boundary: a correct reference doesn't make withdrawn evidence reliable.
If your assignment concerns publication ethics, a retracted paper may be part of the material you analyze. Explain why it remains in the discussion and credit the notice for statements about its status. If you're writing a literature review to establish what research supports a conclusion, reassess the affected evidence instead. Don't silently replace the citation while keeping a claim that no other source supports.
Verify the current status using identifiers
Open the publisher page and its linked notice
Start with the original article's DOI or publisher landing page. Match the title, author information, journal, and publication details to the item in your draft. A similar title isn't enough. An old downloaded PDF may not show a later update.
Open the linked notice and read its actual explanation. Record the notice's date and identifier separately from the original article's. Check whether the notice concerns the entire paper, a correction to one element, or an unresolved concern. If the notice doesn't explain something, don't fill the gap with a rumor or a conclusion about an author's intentions.
The ICMJE guidance on retractions calls for links in both directions between the original article and the notice, clear retraction labels, and an explanation of the reason. That describes how journals should document the change. If you can't find the notice, ask a librarian or the journal for help rather than treating the missing link as evidence that nothing happened.
If you've only encountered the study in another writer's summary, retrieve the original source first. Check its current status before repeating the summary's interpretation.
Check PubMed where applicable and Crossref metadata
For a paper indexed in PubMed, open its record and inspect linked updates. NLM's PubMed policy distinguishes the record for a retracted publication from the record for its retraction notice. The record labels help identify which item you're viewing; the notice supplies the explanation you need to read.
Crossref's Retraction Watch documentation describes access through its REST API and downloadable dataset. Search using the article identifier where available, then match the result to the publisher's record. Its data also includes some corrections and expressions of concern, but Crossref says those categories are less comprehensive than its retraction data.
No match is not proof of no retraction. A database may not cover the article, an update may not appear in the record you're using, or you may have searched the wrong identifier. Keep the publisher's notice in your verification trail. If sources disagree, record the disagreement and investigate it before relying on the finding.
Separate corrections, expressions of concern and retractions
Use this status/action table as an editing checklist, not as a substitute for reading the notice. Your action depends on the specific finding your paragraph uses.
| Status found | Question to answer | Action in your draft |
|---|---|---|
| Correction or erratum | What changed, and does it affect the information I use? | Read the correction and current article. Revise any affected figure, wording, or interpretation. Don't describe a correction as a retraction. |
| Expression of concern | What concern has the editor identified, and is there a later outcome? | Check for subsequent updates. Explain the uncertainty if discussing the article; don't describe an unresolved concern as a completed retraction. |
| Retraction | Why was the paper retracted, and why am I still mentioning it? | Remove reliance on withdrawn findings. If discussing the retraction, identify the status and cite the relevant article and notice. |
| No update found | Did I check the exact article and the publisher's current record? | Record the places checked and date. Don't call the source certified or assume a search proves that no update exists. |
Don't turn a status label into an allegation. NLM's policy explains that its retraction labels don't distinguish honest error from misconduct or plagiarism. Describe the reason the notice actually gives, with attribution. A retraction alone doesn't tell you every detail about why a research project went wrong.
A correction can also be consequential. If a corrected value changes the comparison in your paragraph, update that comparison even though the article wasn't retracted. If the notice only corrects an author's name, that calls for a different repair. Read before deciding.
Cite the article and notice as distinct objects
Explain the status in your prose
Readers shouldn't have to discover the status by following a DOI after they've accepted your argument. State it near the discussion: name the article, identify it as retracted, and attribute the explanation to the notice. Distinguish what the authors originally claimed from what the journal later reported.
The article and notice may have different authors, dates, titles, and identifiers. Don't assign the original article's DOI to the notice or use the notice's date as if it were the study's publication year. When describing the reason for retraction, use the notice as the source of that information.
Keep source credit in your paper close to the statement it supports. One citation after a paragraph containing the original findings, the notice's explanation, and your analysis can blur three different contributions.
Check the style authority before adding a reference annotation
Use the citation style your assignment or journal requires. Look for its guidance on publication updates, then check any local submission instructions. Don't assume that a bracketed status label is a universal APA, MLA, or Chicago requirement.
NLM's Citing Medicine retraction guidance documents notes connecting an article to a later retraction and a notice to the article it retracts. Those are NLM conventions. They shouldn't be relabeled as APA rules or copied into another style without checking that style's authority.
You can format a supported reference after checking its metadata. That doesn't establish whether the paper is reliable or automatically supply a required status annotation. This guide deliberately avoids a multistyle reference template: the argument-repair workflow applies across styles, but the final reference format needs its own check.
Repair the paragraph, not just the bibliography
Fictional worked example: Every study, result, notice, date, and identifier in this example is invented for teaching. STUDY-A and NOTICE-A are local worksheet labels, not DOIs, PubMed identifiers, or real references.
Before checking: “Weekly planning sessions improve student retention, as STUDY-A demonstrates. The program should therefore become a required part of first-year advising.”
In this fictional case, NOTICE-A says STUDY-A was retracted because the underlying data couldn't support its reported results. The draft uses those results to justify a program recommendation. Adding “retracted” beside the bibliography entry would leave the unsupported recommendation intact.
Revised paragraph for a publication-ethics discussion: “STUDY-A originally reported a retention benefit from weekly planning sessions. NOTICE-A later retracted the paper because its data couldn't support the reported results. The original finding therefore doesn't establish the benefit asserted in this discussion.”
This revised version reports publication history. It doesn't replace the lost evidence for the advising recommendation. If the assignment instead asks which advising program works, remove that unsupported recommendation or find and read independent evidence that directly addresses it. A new source about student satisfaction won't automatically support a claim about retention.
Search the whole draft for the affected study. Its influence may extend beyond the citation-bearing sentence into your introduction, abstract, table, or conclusion. In a literature review, also check whether a review you cite relies on it. Use the workflow to audit literature review sources without assuming that every paper mentioning the retracted study has the same problem.
Keep a verification log you can reproduce
This filled teaching log belongs to the fictional example above. For a real source, replace each field with details you actually checked. Keep a separate entry for each affected article.
| Field | Fictional teaching entry |
|---|---|
| Article identifier | STUDY-A, local fictional worksheet label; no DOI assigned. |
| Status source | NOTICE-A, fictional publisher notice linked to STUDY-A. |
| Notice date | September 20, 2026, invented for this example. |
| Affected claim | Planning sessions improve retention and justify a required advising program. |
| Replacement evidence | None verified. Recommendation removed pending an independent evidence search. |
| Checked date | October 8, 2026, fictional worksheet check. |
In your real log, save the actual article and notice URLs, their identifiers, and enough notes to locate the relevant passage. Record whether you removed, narrowed, or replaced the claim. “Reference fixed” is too vague to show what changed.
If the assignment was already submitted
Contact the professor rather than quietly changing a file that may no longer be the assessed version. Explain which source changed status, when you discovered it, and which passage needs repair. Ask how to submit the correction through the course's process.
Message template: “I discovered that the article identified by [article DOI or publisher URL] has a retraction notice at [notice URL]. My submitted paper uses it in [section or paragraph] to support [claim]. I found the notice on [date]. I propose [removing or revising the claim], and I've attached the affected passage. Please let me know how to provide a corrected version.”
Keep the original submission and your revision record. Don't speculate about penalties or promise that disclosure will produce a particular outcome. Course procedures and the facts of the situation determine what happens next.
Finish source verification before a text check
Before submitting, confirm that each retained claim has support from a source you've read, that publication updates are explained where relevant, and that the article and notice have separate accurate records.
If you also want to review matching text, treat that as a separate task. A similarity scan cannot verify research reliability, resolve a retraction, or supply replacement evidence. You can start a text check after the source audit, but a low similarity percentage doesn't validate the argument.
FAQ
Yes, when the article is relevant to discussing the retraction, publication history, or the research record. Identify its status clearly. Don't present its withdrawn findings as dependable support for a factual claim.
If you describe why or when the paper was retracted, cite the notice that supports that description. Treat the original paper and its notice as distinct publications, and follow your assigned style or journal instructions.
No. A correction and a retraction are different publication updates. Read the correction, determine whether it changes the result you use, and update your claim and source record accordingly.
No. PubMed has a defined indexing scope, and Crossref metadata is a checking aid rather than a guarantee of complete coverage. A missing result doesn't prove that an article has no retraction or other update. Check the publisher page and notice.
Contact your professor with the article identifier, the notice, the affected passage, and your proposed correction. Explain when you discovered the status change and ask how to provide a revised version. Don't assume a particular academic outcome.