Ask two doctoral candidates how long a thesis is and you'll get two wildly different numbers, both correct. A chemistry thesis of 45,000 words and a history thesis of 95,000 words can pass on the same day, in the same country, at the same university. That's because thesis length isn't one convention — it's several, stacked on top of each other by country, discipline and institution. Here is the full stack, and how to find the number that applies to you.
The short answer: 70,000–100,000 words
Across systems and disciplines, most PhD theses land between 70,000 and 100,000 words. Treat that as the center of the distribution, not a rule: plenty of excellent theses pass at 50,000, and in some fields anything over 80,000 raises eyebrows rather than respect. The binding constraint is almost never "how long should it be?" — it's "what is my university's maximum, and what counts toward it?" Those two questions have precise answers, and they're the ones that matter at submission.
For scale, 80,000 words is roughly 280–320 pages of double-spaced text, or about 160–180 pages single-spaced — somewhere between a long trade nonfiction book and a full novel, and comfortably longer than most novels ever published.
By country: three different systems
The biggest single factor in your word limit is where you're enrolled:
| System | Typical limit | How it's enforced |
|---|---|---|
| UK & Ireland | 80,000–100,000 words | Strict, set by university regulations; footnotes typically count |
| Australia & New Zealand | 80,000–100,000 words | Similar to the UK; some universities go lower for lab-based degrees |
| United States | ~150–300 pages | Usually specified in pages, not words, and loosely enforced |
In the UK, the limit is a genuine legal-style constraint: most universities set it at typically 80,000–100,000 words including footnotes, but excluding the bibliography, appendices and usually the abstract. Examiners and administrators actually count, and a thesis over the limit without written permission is a problem you have to fix before it can be examined. Australia and New Zealand broadly mirror the UK approach, with similar caps and similar exclusions.
The United States works differently. US dissertations are usually specified in pages — roughly 150–300 — rather than words, and the number is treated more as guidance than a ceiling. The practical word-count equivalent is about 40,000–90,000 words depending on spacing, figures and how much of the page budget the tables consume.
By discipline: humanities vs STEM
Within any one country, discipline does most of the remaining work:
- Humanities and social sciences (history, literature, politics, anthropology) typically run 80,000–100,000 words. The argument lives in the prose, chapters are long, and the literature review is a substantial document in itself.
- STEM and lab-based sciences commonly run 40,000–80,000 words. Results live in figures, tables, equations and published papers, and many programs allow or require a "thesis by publication" format that keeps the written core compact.
- Arts and practice-based doctorates can be shorter still, because the examined work includes a creative artifact and the written component may be capped at 30,000–50,000 words.
None of these are quality judgments. A 45,000-word materials science thesis containing three published papers is not "less of a PhD" than a 95,000-word historical study — the disciplines are simply measuring different things.
What counts — and what doesn't
The exclusions can swing your effective budget by 15,000 words or more, so check your own regulations rather than borrowing another university's. The common pattern:
- Usually excluded: bibliography and references, appendices (data, code, instruments, transcripts), tables, figures, and often the abstract and acknowledgements.
- Usually included: the main text, and in the UK typically the footnotes as well — a nasty surprise for humanities candidates with discursive notes.
- Varies by institution: whether the abstract, lists of contents/figures, and any front matter count; whether appendices count if they contain argument rather than raw data.
Because the rules differ, the professional move is to find your university's official "degree regulations" or "thesis format" document in your first month, not your last — then track your running total as you draft. Pasting each chapter into the WordlyCount counter gives you the exact included-word figure in seconds, which is far less painful than discovering a 6,000-word overage during final formatting.
Going over: overage permissions exist
If you're over the limit, you are not automatically doomed. Most UK-style universities have a formal process to request an extension to the word limit — typically up to 10% — with a short written case approved by your supervisor and the graduate office. Legitimate reasons include extensive necessary primary-source quotation, a large corpus of analyzed material, or examiners' documented requests during monitoring. What doesn't work is simply hoping nobody counts.
The better path is almost always cutting. Examined prose benefits from the same discipline as any other writing: a chapter trimmed by 15% usually reads better, not thinner. If you need a method for that surgery, our guide to cutting your word count is written for exactly this situation — and the same skills that trim a 12,000-word chapter to 9,000 will serve the whole thesis.
Working to a limit without panic
A useful final framing: nobody ever failed a PhD for being 5,000 words under the limit, and people create real administrative trouble by being 5,000 over it. Write toward the argument, track the count honestly, and let the limit discipline the editing passes rather than the drafting. Candidates who check their total monthly — not nightly — finish with fewer surprises and considerably less anxiety than those who guess.
For comparison with other long-form targets, the same counting discipline applies to a full-length novel or even a structured long blog post — the numbers just have an extra zero or two on the end.