How to Use Wikipedia for Academic Research

Quick Answer: Start on Wikipedia, Cite What It Cites

Most instructors will not accept Wikipedia as a cited source for a research paper. That is not a mysterious campus rule. Wikipedia classifies itself with other encyclopedias: a tertiary source that summarizes published work. The job of an encyclopedia is to introduce a topic. The job of your paper is to use the books, articles, data, and documents underneath that summary.

This URL is the Wikipedia workflow. It is not how to tell if a source is credible in general, not the CRAAP walkthrough, and not the scholarly vs popular type map. Those pages decide whether a published item is good enough. This page decides what to do with the encyclopedia tab you already have open.

What Wikipedia Itself Tells Researchers

You do not need a third-party blog to learn the house rule. Wikipedia:Citing Wikipedia is the project’s how-to page for people who want to put a Wikipedia URL in a paper:

The same page treats that advice as encyclopedia advice, not a Wikipedia-only ban:

The companion how-to, Wikipedia:Researching with Wikipedia, is even blunter about reliability:

And it tells you what to do with the footnotes — the step most students skip:

That last sentence is the whole method. A blue superscript is a claim that a source exists, not proof that the source says what the sentence says. Editors mis-cite. Footnotes wander. Sometimes the cited page is paywalled, off-point, or a different edition. Open it.

Why Wikipedia Is a Weak Thing to Cite

Four properties make Wikipedia a poor citation for most student papers. None of them requires you to pretend the site is useless.

1. Anyone can edit the page you just read

The general disclaimer is not fine print. It is the legal and editorial premise of the project:

Blatant vandalism is often reversed quickly. Subtle errors, one-sided sourcing, and quiet insertions about living people can sit for a long time. The 2005 Seigenthaler biography incident — a hoax that falsely connected journalist John Seigenthaler to the assassinations of John and Robert Kennedy — is why Wikipedia later tightened biographies of living people. The incident is also why “I read it on Wikipedia yesterday” is not a stable claim.

2. There is no formal peer review

Volunteer editors watch Recent Changes, and Featured Articles go through a public review. That is not the same thing as peer review at a journal. The disclaimer says so in those words:

3. It is an encyclopedia, not the study

Wikipedia:Researching with Wikipedia puts the academic consequence in one sentence:

Citing Wikipedia for a finding that already lives in a journal article is citing the map instead of the territory. If the map is right, the article is the better citation. If the map is wrong, you just copied the error.

4. Wikipedia does not treat open wikis as reliable sources either

The core content policy Wikipedia:Verifiability tells editors what they may use. Open wikis sit with blogs and social posts on the “usually not reliable” list:

That is Wikipedia telling Wikipedia not to cite Wikipedia. Your instructor is applying the same standard to you.

What APA 7, MLA 9, and CMOS 18 Actually Say

The old invented sentence — “Wikipedia is not considered a credible source,” pinned on APA 7 — is not in the manual. APA gives a format, then puts the academic decision back on the instructor. Covered in the Publication Manual of the American Psychological Association (7th ed.), §10.3, and reprinted on campus APA guides from the APA Style Wikipedia example:

Two notes on that wording. First, “ask your professor” is the rule; the format is only for the cases where the answer is yes. Second, APA calls Wikipedia a secondary source because it reports other sources. Librarians and Wikipedia itself usually call an encyclopedia tertiary. Do not get stuck on the label. Both descriptions mean the same practical thing: you are looking at a summary, not at the study.

If you are allowed to cite the entry, APA 7 wants a specific revision, not today’s live URL. The Publication Manual (7th ed., p. 329) is explicit:

The official worked example on the APA Style site is the 8 December 2019 revision of “Oil painting.” That oldid URL is the archive, so the official example has no “Retrieved …” date:

Wikipedia:Citing Wikipedia still prints an APA 5th-edition sample for the 22 July 2004 revision of “Plagiarism,” with n.d. and a retrieval date. Do not copy that sample into an APA 7 paper. Use the Oil painting pattern: revision date in parentheses, Wikipedia as the italic container, and the oldid URL.

MLA 9 treats a Wikipedia article as an encyclopedia entry in a web container: article title in quotation marks, Wikipedia as the container, Wikimedia Foundation as publisher, the revision date, and a permalink. Wikipedia’s citing page still shows an older (MLA 8) sample built on that same 2004 “Plagiarism” revision. Keep Wikipedia’s string as Wikipedia’s illustration, then apply the MLA 9 container to the same verified oldid:

Chicago 18 puts well-known encyclopedias in notes, not in the bibliography. Douglas College’s CMOS 18 reprint of §14.130 is the current rule:

For a live wiki, CMOS 18 §14.131 wants a revision or access date instead of an edition number, and it allows a timestamp. Wikipedia does not have a print counterpart, so the title is not italicized (CMOS 18 §§14.103–14.104, as Douglas College notes). Wikipedia’s citing page still shows the older s.v. form. Prefer the CMOS 18 note on the same 2004 revision:

Style manuals tell you how to cite an encyclopedia when you must. They do not overrule a syllabus that says “no Wikipedia.” If you have permission and need a formatted entry, use the Wikipedia citation generator — then still prefer the underlying source.

A Real Walkthrough: Wikipedia’s Cerebellum Article

Do not invent a “classical conditioning → Pavlov” story. Open a real article that Wikipedia itself uses as an example. Wikipedia:Citing Wikipedia lists articles that were later published in peer-reviewed journals. One of them is Cerebellum. The live article banner (opened for this refresh) states that the page was submitted to WikiJournal of Medicine for external academic peer review in 2016 and that the version of record is:

Wikipedia’s own rule for this case is sharp: cite the published paper, not the live wiki page, because the Wikipedia article can diverge after review. The student move is not “cite Wikipedia, Cerebellum.” It is: open the DOI, read Wright, Skaggs, and Nielsen, and cite that article if that paper is what supports your sentence.

Three failure modes show up on almost every article, including good ones:

• The footnote does not support the sentence. Wikipedia said X; the cited page is about Y, or only about a nearby claim. Believe the source, not the superscript.
• *The footnote is the wrong kind of source.* A BBC News item, a personal page, or another encyclopedia is not suddenly scholarly because a Wikipedia editor used it.
• You needed a neighboring paper, not this one. If your claim is about Watson and Crick’s 1953 Nature letter, find Watson and Crick 1953 and read that. Do not cite Wright 2016 — or a nearby Franklin crystallography paper — just because it was easier to copy.

Harvard’s Guide to Using Sources makes the same research rule without dressing it up:

Copies of Wikipedia Are Not Confirmation

A second website that “agrees” with Wikipedia is often just Wikipedia with the license notice stripped. Wikipedia:Researching with Wikipedia names the trap:

If the confirming page repeats Wikipedia’s section order, its leftover wiki-markup, or its exact wording, you have one source, not two. Open a book, a journal, or an official report instead.

How to Judge a Wikipedia Article Before You Trust the Footnotes

Some articles are careful surveys. Some are stubs, fansites, or battlefields. You are allowed to use quality signals to decide how hard to work the footnotes. You are not allowed to treat a bronze star as a permission slip to cite the page.

Featured Articles and Good Articles

A bronze star in the upper right means Featured Article; a green plus means Good Article. Those badges are real community reviews. They are still reviews by volunteer editors, and the article can change after the badge is awarded. Wikipedia:Researching with Wikipedia notes that as of October 2024, 1,694 one-time featured articles had later lost that status, only 73 of those had regained it, and there were 6,608 current featured articles. The featured-article criteria themselves describe a survey of literature, not a primary source:

Read that second line as a gift: a good Featured Article already did the bibliography work. Steal the bibliography. Do not steal the summary.

Warning banners, Talk, and View history

Before you harvest footnotes, look at the chrome around the article:

• Maintenance banners at the top (“The neutrality of this article is disputed,” “This article needs additional citations”) are not decorations. They tell you the page is unstable or thin.
• [citation needed] after a sentence means even Wikipedia’s editors would not let you rely on that line.
• Talk is where editors argue about wording, sources, and weight. A long, bitter Talk page is a reason to slow down, especially on living people and current events.
• View history shows who changed what, and when. A quiet science article with years of small sourced edits is a different object from a page that swung 800 words yesterday.
• Permanent link (under Tools) is the oldid URL APA 7 wants if you must cite the page.

The Accuracy Question: Nature (2005), Not a Vibes Test

Students hear two slogans: “Wikipedia is always wrong” and “Wikipedia is as good as Britannica, so I can cite it.” Neither is the finding. In December 2005, Nature asked expert reviewers to compare science entries from Wikipedia and Encyclopaedia Britannica. Jim Giles reported the result in Internet encyclopaedias go head to head (Nature, 438, 900–901). The standfirst is the sentence people remember:

The comparison itself, among 42 reviewed pairs, was not a tie and not a wipeout:

Encyclopaedia Britannica disputed the method in 2006. Nature stood by the story. You do not need to referee that argument to use the result: some Wikipedia science articles were close to a respected encyclopedia, and both encyclopedias still contained errors. Closeness on a science-accuracy test does not make either work a substitute for the journal article you should be reading. It also says nothing about biographies, politics, medicine for living patients, or breaking news — the places Wikipedia is most volatile.

Health Papers Still Cite Wikipedia. That Is Not Permission.

In March 2014, The BMJ published a bibliometric review of Wikipedia citations in indexed health-science journals: Bould, Hladkowicz, Pigford, Ufholz, Postonogova, Shin, and Boet, References that anyone can edit (BMJ, 348, Article g1585). They recovered 2,049 Wikipedia citations in 1,433 articles from 1,008 journals indexed in Medline, PubMed, or Embase. More than half of those citations were definitions or descriptions. The citations were not confined to low-impact titles. The authors’ conclusion is the line that matters for a student paper:

A Nature or BMJ article that cites Wikipedia for a definition does not rewrite your syllabus. Wikipedia’s own disclaimer also says the site is not professional medical, legal, or financial advice. For clinical or policy claims, follow the published guideline or trial, not the encyclopedia summary. See Medical citation guide when the source you actually need is AMA, ICMJE, or NLM.

When Citing Wikipedia Is Actually the Right Move

Wikipedia:Citing Wikipedia lists the exceptions that matter. They are narrower than “my professor is chill.”

If you are writing about Wikipedia’s notability guideline, the growth of a language edition, an edit war, or how a Talk page framed a controversy, the wiki page is a primary source for that event. Cite a specific oldid. Quote sparingly. Say what the page was doing, not what you wish it had proved about the world.

The same help page allows a second, weaker case: an obscure topic where the Wikipedia article is the best compiled account you can reach, and you have already told your instructor that the published record is thin. Wikipedia’s own example is a sourced article such as Siege of Compiègne. “This siege is on Wikipedia and nowhere in our library” is a conversation. It is not a default setting.

Copying Wikipedia prose without a citation is still plagiarism, even if you agree the page is a bad source. The citing guide is explicit: if you quote or paraphrase the article, you must attribute it, or you have both an academic-integrity problem and a CC BY-SA license problem.

Better Encyclopedias When You Still Need an Overview

Sometimes you do need a signed, citable overview — a concept you have never seen, a field’s vocabulary, a first map of a debate. Use an encyclopedia with named authors and a publisher your library already pays for.

APA 7’s own encyclopedia example — not Wikipedia — is George Graham’s signed Stanford Encyclopedia of Philosophy entry, the Publication Manual pattern on p. 328 as campus APA guides reprint it:

See the difference: a named author, an editor, an edition you can recover, a university publisher. That is what “cite an encyclopedia” looks like when the encyclopedia is built for citation.

A Workflow You Can Finish in One Sitting

Keep Wikipedia in your research notes if it helped you. That is using a tool. Putting it in the reference list of a history seminar paper is presenting the tool as evidence.

Common Misconceptions

“Wikipedia is always wrong”

No. Giles (2005) is the reason this slogan died for science articles, and many technical pages are carefully sourced. The problem for a paper is not that every sentence is false. The problem is that you cannot see who is accountable, the text can move, and the page is still an encyclopedia.

“Using Wikipedia at all is cheating”

No. Reading it for orientation is normal. Professors do it. Wikipedia’s own research help recommends it. Cheating is copying the article, or citing the summary instead of the source you should have opened.

“A Featured Article can go in the reference list”

No. The badge says the survey is stronger than average. It does not change the genre. Cite the literature the article surveyed.

“If Wikipedia cites it, the source is reliable”

No. Verifiability is a rule for editors, not a guarantee that every footnote you meet is a scholarly source. Evaluate the cited item on its own, the way you would if you had found it in a database. Use How to evaluate sources when the item is unfamiliar.

“Other wikis are fine if Wikipedia is not”

Fandom wikis, course wikis, and most specialist wikis have weaker sourcing norms than English Wikipedia. They can be primary sources if you are studying a fandom or a community. They are not a workaround for a banned encyclopedia.

“CMOS still wants s.v. for Wikipedia”

No. CMOS 18 §14.130 dropped the Latin abbreviation. Wikipedia’s citing page still shows the older s.v. note. Use a CMOS 18 note with a revision date or timestamp, and keep well-known encyclopedias out of the bibliography unless your instructor says otherwise.

Related Guides

• How to Tell If a Source Is Credible
• How to Evaluate Sources for Academic Research
• Scholarly vs Popular Sources
• How to Find Scholarly Sources
• Primary vs Secondary Sources
• How to Avoid Plagiarism