BlogCitApp

Citation guides

How to Cite a PDF: APA, MLA and Harvard

A PDF is a file format, not a source type — so cite what it actually is. How to identify whether your PDF is a journal article, report, book chapter or thesis, where to find the metadata, which page number to use, and what to do with scanned files that have no author.

Ihsan Hossain15 August 2026 8 min read

There is no such thing as "PDF format" in APA, MLA or Harvard. PDF is a container — the same one used for journal articles, government reports, dissertations, book chapters and lecture slides — and every style asks you to cite the work inside it.

So the real question is never "how do I cite a PDF". It is "what is this document, actually?" Answer that and the format follows.

Step 1: identify what you are holding

Open the first page and look for these signals, in order:

  • A journal name, volume and page range in the header → it is a journal article. Cite it as one.
  • A DOI (a string starting 10.) → almost certainly a journal article or a formally published report, and the DOI is the best identifier you can give.
  • An organisation's logo and a report number → a report. The organisation is the author.
  • "A thesis submitted in partial fulfilment…" → a thesis or dissertation.
  • An ISBN and a publisher → a book or a chapter from one.
  • None of the above → treat it as a standalone document or web page.

Where the metadata hides

If the first page is unhelpful, three more places usually have what you need:

  • Document properties. In most readers, File → Properties shows an embedded title, author and creation date. Treat the date with suspicion — it is often the date the file was generated, not published.
  • The last page. Reports frequently put the imprint, publication date and report number in the back matter.
  • The DOI. Paste it into a search and you get the full, authoritative record — which beats anything you can read off the page.

Step 2: apply the format for that source type

A journal article that happens to be a PDF

APA 7

Andersson, K., & Berg, L. (2024). Sleep deprivation and working memory. Journal of Sleep Research, 33(2), 145–162. https://doi.org/10.1111/jsr.13845

MLA 9

Andersson, Karin, and Lars Berg. "Sleep Deprivation and Working Memory." Journal of Sleep Research, vol. 33, no. 2, 2024, pp. 145–62, https://doi.org/10.1111/jsr.13845.

Harvard

Andersson, K. and Berg, L. (2024) ‘Sleep deprivation and working memory’, Journal of Sleep Research, 33(2), pp. 145–162. doi: 10.1111/jsr.13845.

Note what is absent from all three: any mention that you read a PDF. See our full guide to citing journal articles.

A report from an organisation

APA 7

World Health Organization. (2025). Global status report on sleep health (WHO/MSD/2025.3). https://www.who.int/publications/sleep-2025

Harvard

World Health Organization (2025) Global status report on sleep health. WHO/MSD/2025.3. Geneva: WHO. Available at: https://www.who.int/publications/sleep-2025 (Accessed: 15 August 2026).

The report number is worth including whenever there is one — it is the fastest way for a reader to find the exact edition you used.

A thesis or dissertation

APA 7

Ek, P. (2022). Sleep and academic performance in adolescents [Doctoral dissertation, Uppsala University]. DiVA. https://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-123456

Lecture slides — the real exception

This is the one case where a format description belongs, because there is no other way to tell the reader what kind of object it is:

APA 7 — lecture slides

Berg, L. (2026). Circadian rhythms [PDF slides]. BIOL2041: Human Physiology. Uppsala University. https://studium.uu.se/courses/2041/files

If the link only works for enrolled students, say where it lives rather than pretending it is public.

Step 3: get the page number right

This trips up more students than the reference format does. A PDF has two sets of page numbers: the ones printed on the page, and the ones your viewer counts. They are usually different, because covers, blank pages and roman-numbered front matter all shift the count.

The two numbers

PDF viewer says: page 12 of 48 Printed on page: 4 Cite: (Andersson, 2024, p. 4)

Always cite the printed number. It is the one that exists in every copy of the document, including the paper one your marker might pull off a shelf. If there are no printed numbers at all, count from the first page of actual content and note that you are doing so, or cite a section heading instead.

Awkward cases

No author

Work down the ladder — named person, then publishing organisation, then title. Most authorless PDFs are institutional, so the issuing body takes the author position: a ministry, a company, a university. Only when no organisation is identifiable does the title move forward.

No date

Use n.d. in APA and Harvard. Resist the temptation to use the file's creation date from document properties, or a copyright year in the footer — neither tells you when the content was published.

A scanned PDF with no text layer

If you cannot select the text with your cursor, the file is an image and no tool can read it automatically — including ours. You will have to type the details in by hand. It is also worth checking whether a proper digital version exists: scanned copies are often unofficial re-distributions, and the published original is a better source in every respect.

A paywalled PDF

Cite it normally. Your reader does not need to be able to open the link for the reference to be valid — that is what the DOI is for. Never substitute a link to a pirated copy.

Let the file fill in its own details

Most of what you need is already inside the document. CitApp reads a PDF in your browser — the file never leaves your computer — and pulls out the title, authors, year and DOI, then formats the reference in APA, MLA, Harvard or any of 149 built-in and library styles. If you save a quotation from it, CitApp also works out which printed page the passage came from, so the page number in your citation is the one your reader can actually find.

Try the citation generator, or see our guides to citing books and citing websites.

IH

Written by Ihsan Hossain

Founder of CitApp, a citation tool built to make source formatting effortless for students. Guides on this blog are checked against the current official style guides (APA 7, MLA 9, and university Harvard variants).

Frequently asked questions

Do I need to write [PDF] in the citation?

Almost never. APA 7, MLA 9 and Harvard all want you to describe what the work is, not what file it arrived in, so a journal article saved as a PDF is cited exactly like a journal article. The one exception is a document whose format genuinely matters and that has no other category — lecture slides, for example, are given a bracketed description like [PowerPoint slides] or [PDF] in APA so the reader knows what to expect.

How do I cite a PDF with no author?

Move down the ladder: a named person first, then the organisation that published it, then the title. Most authorless PDFs are corporate or government reports, so the issuing body becomes the author — the Department of Health, the World Bank, the university. Only when there is no identifiable organisation does the title move into the author position.

Should I use the PDF page number or the printed page number?

The printed page number — the one shown on the page itself. Your reader needs a number they can find in any copy of the document, and a PDF viewer's counter is often offset by cover pages and front matter, so page 12 of the file may be page 4 of the report. If the document has no printed numbers at all, count from the first page of the actual content and say so, or cite a section heading instead.

How do I cite a PDF from my university's course site?

If the document exists publicly elsewhere — a published article, a public report — cite that original version, because a Canvas or Blackboard link only works for people enrolled in the course. If it is genuinely course material written by your lecturer, cite it as unpublished course material with the instructor as author, the course code, and the institution.