The Vancouver Group and the history of Vancouver style

Vancouver style is named after a meeting. In January 1978 a small group of editors of general medical journals met in Vancouver, British Columbia, agreed how manuscripts sent to their journals should be prepared, and asked the US National Library of Medicine (NLM) to settle the form of references (Huth, 1981). The group became the International Committee of Medical Journal Editors (ICMJE), and the NLM's Citing Medicine now sets the details.

EdCitation publishes this history. It rests on the committee's own printings from 1978 to 1997, read in scans and the ICMJE's archive, the ICMJE's and NLM's current pages, and accounts by two people who were there, Edward Huth and Kathleen Case, all read on 24 September 2026. Where the sources disagree, we say so.

Today's rules are in our Vancouver referencing guide, and EdCitation's free Cite a source sets any published work in Vancouver from the publisher's record.

Where did Vancouver style come from?

From a secretary's complaint, two editors' agreements and a library's formats. Huth and Case (2004) begin in Seattle in 1968, with Augusta Litwer, secretary to the nephrologist Belding Scribner, tired of retyping references whenever a rejected paper went to another journal. Urged by a university librarian, Litwer wrote to the editors of Annals of Internal Medicine, JAMA and the New England Journal of Medicine to ask why they could not share one format.

The first agreement, 1970

Those editors and others agreed in 1970 to use the Index Medicus formats set by the NLM (Huth & Case, 2004). An unsigned BMJ editorial of May 1978 tells the story with other numbers: the secretary wrote to six journals, asking why each needed its own "little pet form of references", and editors of 30 journals agreed ("Declaration of Vancouver," 1978). Huth and Case count eighteen that signed. The editorial also gives the scale of the problem: a speaker at a 1977 conference claimed a survey had found 33 reference styles in 52 journals.

The Vancouver meeting, January 1978

In 1976 Huth, editor of Annals, and Stephen Lock, editor of the British Medical Journal, discussed an international agreement on references. Huth, John Murray and Therese Southgate of JAMA organised a meeting, and Lock proposed neutral ground between Britain and the United States (Huth & Case, 2004). References were the main and most quarrelsome business: Huth wanted the American National Standards Institute (ANSI) standard, others objected, and the group, calling itself the International Steering Committee, left the formats to the NLM.

The agreement covered manuscripts, not printed pages: journals would accept papers prepared this way and keep their own house style in print (Huth, 1979).

Why is it called Vancouver style?

Because of that first meeting place, and the editors used the name before universities did. The 1979 printing calls the 1978 text the original Vancouver document (International Steering Committee of Medical Editors, 1979). Huth (1979) calls it the Vancouver Declaration and writes of the "NLM (Vancouver) style". By January 1981 the BMJ listed journals that had agreed to receive manuscripts "in the Vancouver style" (Huth, 1981), and the 1988 and 1991 editions open by calling themselves that.

The committee itself was the Vancouver group. Huth's report of its October 1980 meeting at Leeds Castle, in Kent, already calls it the International Committee of Medical Journal Editors, with more than 130 journals in the agreement (Huth, 1981). Its membership page still says it was once the Vancouver Group (International Committee of Medical Journal Editors [ICMJE], n.d.-b), but neither its references section nor the NLM's samples call the format Vancouver.

Were the Uniform Requirements first published in 1978 or 1979?

Both years are right, for different texts. The BMJ printed a draft on 20 May 1978, headed "For Debate", with a tentative list of 19 participating journals (International Steering Committee of Medical Editors, 1978). A second meeting, in Montreal on 5 February 1979, approved it with minor amendments, and the agreed version appeared early in 1979, in Annals in January and the BMJ on 24 February (International Steering Committee of Medical Editors, 1979).

The ICMJE's own editions changed their minds. Those of 1997 and 2007 said the requirements were first published in 1979; the August 2013 edition says 1978, as the history page does now (ICMJE, 1997, 2007, 2013, n.d.-a). Huth and Case (2004) split the difference: conceived in 1978, born in 1979.

What did each edition change about references?

First the order of the journal reference, then what it covered, then who held the rules. Each row comes from that edition, except 2003, from Huth and Case (2004).

1978 to 1979: the year moves forward

The draft's example of a correct journal reference ended "N Engl J Med 294: 687-690, 1976". The 1979 text gave the same paper as "Soter NA, Wasserman SI, Austen KF. ... N Engl J Med 1976;294:687-90." In between, the NLM had chosen between the two full forms ANSI's 1977 standard gave for a journal article, taking the one with the date before the volume, an order tied, Huth (1979) explains, to "how the bound volumes of a journal sit on a library shelf". Index Medicus was to follow in January 1980, and its notice of the change is dated that month (Huth, 1981).

The draft's own example needed checking. In the scan we read, it has "eosinophilic" where the paper's title has "eosinophil", and spells two authors Solter and Auster; the 1979 version has them right. The rule above it changed too: references "must be verified against the original documents" became verified by the author(s).

1991: six authors, and the month and issue

From 1979 to 1988 a reference named every author up to six, and for seven or more only the first three and et al. From 1991 it gave the first six (ICMJE, 1991), the rule the NLM's samples still show. Huth and Case (2004) call this a cut from seven names to six; the printed rules show a rise from three to six. The 1991 example also added the month and issue, "Gastroenterology 1980 Aug;79(2):311-4", optional for a journal paginated through the volume.

1997 to 2013: the internet, and a handover

The 1997 edition cited a journal article online as "[serial online]", with "[cited 1996 Jun 5]" and "Available from: URL:", forerunners of today's "[Internet]" and "[cited]", and noted that the NLM then listed up to 25 authors (ICMJE, 1997). In 2003 the reference examples gave way to a link to the NLM's samples page (Huth & Case, 2004), and in October 2007 the ICMJE pointed authors to Citing Medicine (ICMJE, 2007).

How did the NLM come to set the details?

By the Vancouver decision, and by keeping its formats current since. Citing Medicine, second edition, online from 2007 and last updated on 2 October 2015, replaces the NLM's 1991 recommended formats and their 2001 supplement for internet sources (Patrias, 2007). It draws on the NLM's MEDLARS indexing manual, the NISO standard Z39.29-2005 that succeeded the 1977 ANSI standard, and ISO 690, and where they conflict the NLM's manual wins. Its Appendix F lists where MEDLINE's own citations depart from it.

The NLM's samples page, reviewed on 2 June 2026, still sits at an address ending "uniform_requirements.html"; it still stops at six names and et al, while noting that MEDLINE records from the NLM now carry every author (National Library of Medicine, 2026b). PubMed's Cite button offers NLM format beside AMA, MLA and APA, and its user guide warns that no machine capitalises titles correctly every time (National Library of Medicine, 2026a).

What the ICMJE asks of references now

The January 2026 Recommendations keep the 1978 core, numbers in order of first mention and MEDLINE's abbreviations, and add three duties: mark a preprint as one, never cite AI output as a primary source, and check in PubMed that no reference is retracted (ICMJE, 2026). Our guide to retracted papers in your reference list covers the last.

What could not be confirmed?

Five details, where the sources disagree or were out of reach.

• Who was at the meeting. No source lists the people in the room, or the days in January. The 1978 draft names 11 committee members, the 1979 text 12, adding Robert W Mayo, and Huth and Case (2004) mark those 12 as founders. Connor (2013) calls CMAJ a founding member; it is on the 1978 list of journals, but Huth and Case name no CMAJ editor among the founders.
• Three journals or four. Huth (1979) names four that had printed the agreed text early in 1979; Huth (1981) and the 1982 edition say three.
• The 1970 numbers. Six journals or three, and 30 or 18, as above. Huth and Case cite a New York Times report of January 1970 that we could not read.
• Citing Medicine's first edition. We found no date for it. The NCBI Bookshelf blocked automated reading with a browser check, so we read archived copies from 2025.
• The first "Vancouver style". The earliest use we found in print is from January 1981; an earlier one may exist.

How do I cite the Vancouver Group's own documents in Vancouver?

From the record of the printing you read, checked against that printing. Our runs, of 24 September 2026, are quoted as they came back.

1. Find the printing. EdCitation's Find sources searches about 300 million published works. The requirements' full title, most cited first, returned 18,058 records, with printings from 1978 to 2010; a search for the 1978 agreement put Huth's 1981 report first.
2. Build the entry in Cite a source, from the DOI, the PubMed ID, the ISBN, the title or the web address.
3. Check the author, capitals and date against the printing.
4. Check the journal's name at the time: abbreviations follow the title it had then.
5. Check the finished list in Verify references.

The 1979 requirements, from the Annals DOI

Cite a source took the DOI 10.7326/0003-4819-90-1-95 to Crossref's record, reported no retraction, and gave:

International Steering Committee. Uniform Requirements for Manuscripts Submitted to Biomedical Journals. Ann Intern Med. 1979 Jan 1;90(1):95-9. doi: 10.7326/0003-4819-90-1-95.

The abbreviation, volume, issue, pages and the committee's name are right. Two things from the publisher's record need changing: the title case (the NLM uses sentence case), and the day, since MEDLINE's record gives only 1979 Jan. Corrected:

International Steering Committee. Uniform requirements for manuscripts submitted to biomedical journals. Ann Intern Med. 1979 Jan;90(1):95-9. doi: 10.7326/0003-4819-90-1-95.

Huth's 1979 article, from its DOI

Huth E. New forms for references. BMJ. 1979 Jun 23;1(6179):1697-8. doi: 10.1136/bmj.1.6179.1697.

In 1979 the journal was the British Medical Journal, Br Med J to MEDLINE and to the article's own reference list; Cite a source printed today's name. Change BMJ to Br Med J.

A preprint, and the DOI's link

For the arXiv preprint the NLM uses as a sample, Cite a source marked the Vancouver entry "[Preprint]", added "[cited 2026 Sep 24]", and gave the DOI's own link after "Available from:". The NLM's sample adds the arXiv number and version, arXiv:1702.01006v1, if your journal wants it.

Which EdCitation tools help with Vancouver references?

Cite a source is the best tool for a Vancouver reference, because every entry it gives is taken from the record a publisher deposited, never composed, and a chatbot generating references from patterns cannot promise that. The committee's draft shows why: its model of a correct reference reached print with a wrong word and, as we read the scan, two misspelt names. Paste a PubMed ID, DOI, ISBN, title or web address; the entry comes back in Vancouver, or in Harvard, IEEE, APA 7, MLA 9 or Chicago 18 author-date, with the source screened for retraction. It is free, with no account.

Verify references on three Vancouver entries

We gave Verify references, also free, three entries: Huth's 1979 article with the period abbreviation Br Med J, his 1981 report dated 1980, and one we invented, with a made-up author, title and journal.

• The 1979 article, verified: "Title, year and first author match a published record."
• The misdated report, "check this": "The work was published in 1981, not 1980; the record has no date in 1980." Its record line named Annals of Internal Medicine, where Crossref records Huth's article under the same title in January 1981. Keep your entry's BMJ details.
• The invented entry, not found: "No publisher's or registry's record matches this reference."

A retracted paper would have been flagged, and a failed lookup is reported as could not check, not as not found.

Where to look twice

The Uniform Requirements were printed in several journals under one title, and a check by title can land on a sibling. An entry with the 1978 BMJ volume and pages but the year 1979 came back verified, matched to the 1979 Annals printing. The correct 1978 entry came back verified, Br Med J accepted as the journal's name then. Read the record line against your entry. Huth and Case's article came back not found: it has no DOI, and Crossref has no record of it, so hold such an entry against the ICMJE's copy.

A free account keeps the list, which downloads as Word, BibTeX or RIS. References from a file takes the whole manuscript and pairs every numbered citation with its entry: three checks a month with a free account, and no limit with Pro at $8 a month (pricing). EdCitation never writes any part of a paper.

Quick questions

Why is it called Vancouver style?

Because the editors who agreed it first met in Vancouver, British Columbia, in January 1978, and called the result the Vancouver document.

Who created Vancouver style?

A small group of general medical journal editors, later the ICMJE, with the US National Library of Medicine, which they asked to set the reference formats.

Is Vancouver style the same as NLM style?

For references, largely yes: the ICMJE's Recommendations send authors to the NLM's samples and Citing Medicine, neither of which calls the style Vancouver.

When were the Uniform Requirements first published?

A draft appeared in the BMJ on 20 May 1978; the agreed version followed early in 1979, after a meeting in Montreal.

Can EdCitation build Vancouver references?

Yes. EdCitation's free Cite a source builds a Vancouver entry from a DOI, PubMed ID, ISBN, title or web address, from the publisher's record, and screens it for retraction.

References

• Connor, J. J. (2013). CMAJ instruction of authors in the 20th century: From literary style to Vancouver style. CMAJ, 185(5), 419–420. https://doi.org/10.1503/cmaj.111430
• Declaration of Vancouver. (1978). British Medical Journal, 1(6123), 1302–1303. https://doi.org/10.1136/bmj.1.6123.1302
• Huth, E. (1979). New forms for references. British Medical Journal, 1(6179), 1697–1698. https://doi.org/10.1136/bmj.1.6179.1697
• Huth, E. (1981). Manuscript requirements: The advance from Vancouver. British Medical Journal, 282(6257), 55–56. https://doi.org/10.1136/bmj.282.6257.55
• Huth, E. J., & Case, K. (2004). The URM: Twenty-five years old. Science Editor, 27(1), 17–21. https://www.icmje.org/recommendations/archives/summary78-04.pdf
• International Committee of Medical Journal Editors. (n.d.-a). History of the recommendations. https://www.icmje.org/recommendations/browse/about-the-recommendations/history-of-the-recommendations.html
• International Committee of Medical Journal Editors. (n.d.-b). ICMJE membership. https://www.icmje.org/about-icmje/faqs/icmje-membership/
• International Committee of Medical Journal Editors. (1982). Uniform requirements for manuscripts submitted to biomedical journals. British Medical Journal, 284(6331), 1766–1770. https://doi.org/10.1136/bmj.284.6331.1766
• International Committee of Medical Journal Editors. (1988). Uniform requirements for manuscripts submitted to biomedical journals. British Medical Journal, 296(6619), 401–405. https://doi.org/10.1136/bmj.296.6619.401
• International Committee of Medical Journal Editors. (1991). Uniform requirements for manuscripts submitted to biomedical journals. BMJ, 302(6772), 338–341. https://doi.org/10.1136/bmj.302.6772.338
• International Committee of Medical Journal Editors. (1997). Uniform requirements for manuscripts submitted to biomedical journals. New England Journal of Medicine, 336(4), 309–315. https://doi.org/10.1056/NEJM199701233360422
• International Committee of Medical Journal Editors. (2007). Uniform requirements for manuscripts submitted to biomedical journals: Writing and editing for biomedical publication (Updated October 2007). https://www.icmje.org/recommendations/archives/2007_urm.pdf
• International Committee of Medical Journal Editors. (2013). Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals (Updated August 2013). https://www.icmje.org/recommendations/archives/2013_aug_urm.pdf
• International Committee of Medical Journal Editors. (2026). Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals: Preparing a manuscript for submission to a medical journal. https://www.icmje.org/recommendations/browse/manuscript-preparation/preparing-for-submission.html
• International Steering Committee of Medical Editors. (1978). Uniform requirements for manuscripts submitted to biomedical journals. British Medical Journal, 1(6123), 1334–1336. https://doi.org/10.1136/bmj.1.6123.1334
• International Steering Committee of Medical Editors. (1979). Uniform requirements for manuscripts submitted to biomedical journals. British Medical Journal, 1(6162), 532–535. https://doi.org/10.1136/bmj.1.6162.532
• National Library of Medicine. (2026a). PubMed user guide. https://pubmed.ncbi.nlm.nih.gov/help/
• National Library of Medicine. (2026b). Samples of formatted references for authors of journal articles. https://www.nlm.nih.gov/bsd/uniform_requirements.html
• Patrias, K. (2007). Citing medicine: The NLM style guide for authors, editors, and publishers (D. L. Wendling, Tech. Ed.; 2nd ed.). National Library of Medicine. https://www.ncbi.nlm.nih.gov/books/NBK7256/