Ask any group of final-year students how many references a 10,000-word dissertation should have and you will get a confident range of answers, most of them different. One student will swear their supervisor said fifty. Another will insist that anything under a hundred looks thin. A third will tell you there is no number at all and that worrying about it is a waste of time. The frustrating truth is that all three are partly right, because there is no national rule for dissertation reference counts. What exists instead is a set of expectations that shift with your subject, your level of study, the kind of dissertation you are writing, and the way your own department marks. This guide gives you a realistic range for a 10,000-word dissertation, explains why the range is as wide as it is, and shows you how to judge whether your own list is doing its job.
The short answer first, so you can stop worrying and start planning. For a typical 10,000-word undergraduate or taught Masters dissertation, most well-supported examples fall somewhere between sixty and one hundred references in total. Some strong dissertations sit below that range and some sit above it, and neither fact on its own tells you anything about the mark. The number matters far less than what the references are doing: whether each one supports a claim, frames a debate, justifies a method, or helps you interpret a result. A dissertation with eighty references that are genuinely read and used will always beat a dissertation with one hundred and twenty references that were skimmed, stacked, or added to hit a target.
Why there is no official number
Universities rarely publish a required reference count, and there is a good reason for that. A reference count is a proxy for something else: the breadth and depth of your engagement with existing scholarship. A marker can see that engagement directly, in the way you discuss sources, compare them, and use them to build your own argument. Setting a fixed number would encourage exactly the wrong behaviour, which is students hunting for extra sources to reach a quota instead of reading the sources that matter properly. Your dissertation handbook may say something about the quality or type of sources expected, for example that you should draw mainly on peer-reviewed journals, or that textbooks alone will not be enough, but it will almost never say that you must have precisely seventy-five references.
That is why the advice you find online varies so much. Some guides suggest a simple rule of thumb, such as eight to twelve references for every thousand words, which would put a 10,000-word dissertation at eighty to one hundred and twenty. Others suggest a leaner figure, such as one substantial reference every two hundred and fifty to three hundred and fifty words, which lands nearer thirty to forty before any adjustment for the kind of writing a dissertation involves. Both rules are trying to describe the same thing from different angles, and both break down if you apply them mechanically. The first rule overcounts in empirical chapters where your own data does the talking. The second undercounts in a literature review chapter, where surveying the field is the whole point. The honest answer is a range with reasons, not a single number, and this guide works through those reasons chapter by chapter.
What the range looks like by level and subject
Level of study is the single biggest influence on the count. An undergraduate dissertation is expected to show that you can find, understand, and apply the core literature of your topic. A taught Masters dissertation is expected to go further: to engage with primary research, to notice where studies disagree, and to position your own work in relation to an ongoing debate. That difference alone can add twenty or thirty references to a dissertation of the same length, because the Masters student is expected to read more widely before writing. Doctoral work goes further still, but a 10,000-word piece at doctoral level is usually a single chapter or a journal-style paper rather than a whole dissertation, so the comparison is less useful here.
Subject matters almost as much. In the humanities and much of the social sciences, arguments are built by engaging with other writers at length, so dissertations in history, literature, politics, sociology, and education tend towards the upper end of the range. A dissertation that discusses a contested concept across several schools of thought simply needs more sources than one that tests a single hypothesis. In the sciences and in many applied subjects, the literature review is often tighter and more recent, and the results and discussion chapters lean on your own data, so the total can sit lower without any loss of rigour. A laboratory-based dissertation with fifty carefully chosen references can be every bit as strong as a humanities dissertation with ninety, because the work the references are doing is different.
Method matters too. A dissertation based on a systematic or narrative literature review will have the highest count of all, because the review itself is the research: you may screen dozens of studies and cite most of them. A dissertation based on interviews, a survey, or an experiment will usually have fewer references in total, because a large part of the word count is taken up by your method, your findings, and your analysis of data that no one else has produced. A dissertation that develops a theoretical argument, or that compares case studies, will sit somewhere in between. When you read advice that quotes a single number for all 10,000-word dissertations, check whether the writer had your kind of dissertation in mind. Often they did not.
A chapter-by-chapter breakdown
The most practical way to plan your reference count is to think chapter by chapter, because references are not spread evenly through a dissertation. The literature review carries the heaviest load by far, and the conclusion carries almost none. The figures below are planning guides for a typical 10,000-word dissertation, not targets to hit exactly. If your structure differs, move the numbers around to match your own chapters, and keep the principle: references belong where they do work, and the literature review is where most of that work happens.
The introduction, at around a thousand words, usually needs somewhere between five and ten references. Its job is to establish that the topic matters, that a gap or problem exists, and that your research question is worth asking. You do not need to survey the field here, because that is what the next chapter is for. A handful of well-chosen sources that frame the problem, plus perhaps one or two that establish the scale or significance of the issue, is enough. An introduction stuffed with citations reads as a second literature review in miniature and delays the moment the reader learns what you are actually doing.
The literature review, at around three thousand words, is where the bulk of your references live, typically somewhere between thirty and fifty. This is the chapter where you show the breadth of your reading, organise the field into themes or debates, and identify the gap your work addresses. The count here depends on how you write it. A review that discusses twenty studies in depth, comparing their methods and findings, may cite fewer sources than a review that maps a broad field, but it will usually be stronger, because depth of engagement matters more than the number of names dropped. Either way, this chapter alone often accounts for around half of the dissertation’s total references.
The methodology chapter, at around fifteen hundred to two thousand words, usually needs ten to fifteen references. These are different in character from literature review references: they justify your research design, your choice of method, your sampling approach, and your approach to analysis and ethics. Citing a methods textbook or a respected paper for each major decision is what markers look for here. A methodology chapter with no references at all reads as though the design was improvised, while one that cites a source for every sentence reads as padding. Aim for a source behind each significant methodological choice, and no more.
The results and discussion chapters, together around three thousand words, typically carry twenty to thirty references between them. The results section itself may need very few, because it reports your own findings, but the discussion needs sources to interpret those findings: to compare them with previous research, to explain where they agree or disagree, and to draw out implications. A discussion that never returns to the literature reads as disconnected from the field, however interesting the findings are. A discussion that cites a new source for every observation reads as a second literature review. The balance is to return to the key sources from your review, plus a small number of additional sources that help you make sense of something unexpected in your data.
The conclusion, at around a thousand words, needs few references, often between two and five, and sometimes none at all if it is purely a summary of your own argument and findings. New sources should not appear for the first time in a conclusion. If you find yourself wanting to cite something new at the very end, that is usually a sign that the point belongs earlier in the dissertation, in the discussion or the review, rather than being squeezed into the closing pages.
How to tell whether your count is right
Once you have a draft, the count itself is the least useful thing to check. Instead, run through a few tests that markers effectively apply whether they name them or not. First, can you defend every reference? If an examiner asked you why a particular source is in the list, could you say what it contributes in one sentence? If you cannot, the source is probably decoration, and decoration is easy to spot. A reference list padded with sources you have not read is not just wasted effort: it is an academic integrity risk, because you are claiming an engagement with scholarship that did not happen, and an examiner who probes a padded source will find that out quickly.
Second, check the spread. References should appear throughout the body of the dissertation, roughly one working source behind each main point, not stockpiled in the introduction or the literature review and then absent for pages at a time. A long stretch of argument with no citations at all usually means one of two things: either you are presenting your own analysis, which needs no citation, or you are presenting other people’s ideas without acknowledging them, which does. Read through the draft looking for unsupported claims, especially in the discussion, where it is easy to slip from reporting findings into general statements about a field without citing anything.
Third, check the quality and recency of the set as a whole. A list dominated by textbooks, lecture slides, and general websites will look thin even if it is long, because those sources do not demonstrate engagement with current research. A list built mainly on peer-reviewed journal articles and primary research, with a smaller number of books and authoritative reports where they genuinely help, will look strong even if it is shorter. Recency matters differently by subject: in fast-moving fields, a list with nothing from the last five years looks dated, while in history or philosophy, older foundational works are expected and their age is not a weakness. Judge recency against your own field’s pace, not against a universal rule.
Fourth, check for over-citation, which is a real problem even though it is talked about less than under-citation. When citations arrive every sentence or two, your own voice disappears and the dissertation reads as a summary of other people’s work. A useful rule is one to three references in support of each key point: enough to show that the point is grounded, not so many that the reader cannot tell which source actually matters. If you have five sources saying the same thing, keep the strongest one or two and drop the rest. Your argument will be clearer, and your reference count will fall to a level that reflects genuine engagement rather than accumulation.
Planning the number before you write
The best time to think about reference counts is at the start of the project, when the number can guide your reading plan rather than haunting your final week. A practical approach is to set a working range for the dissertation as a whole, then break it down by chapter using the proportions above, and use that breakdown as a reading list target. If your literature review chapter is planned at three thousand words and you expect it to carry around forty references, you know roughly how much reading that chapter requires, and you can spot early whether you are heading for a shortfall. Our free Reference Count Calculator works through the same arithmetic for any word count, level, and assignment type, so you can test a target for your own dissertation before you commit to it.
Be careful, though, about treating any calculator output as a requirement. A planning target is a tool for organising your reading, not a quota to satisfy. If your reading leads you to a tight, well-argued set of sources that falls below the calculator’s range, trust the reading over the calculator, provided each chapter’s argument is properly supported. If your reading genuinely requires more sources than the range suggests, because your topic spans two fields or your review is unusually broad, trust that too. The range describes typical well-supported work. Your dissertation’s job is to support its own claims, and the number that does that is the right number, whether it sits inside the typical range or a little outside it.
Finally, remember the hierarchy that outranks every guide, calculator, and rule of thumb. Your brief comes first: if it specifies a minimum number of sources, a required type of source, or a particular referencing style, follow it exactly. Your supervisor’s guidance comes second, because they know your department’s expectations and your topic’s literature. General advice, including this guide, comes third. If your supervisor tells you that your literature review needs more breadth, that advice is worth more than any number in this article, because it is about your dissertation specifically. Use the range here to plan and to sanity-check your work, and use your brief and your supervisor to make the final judgement.
Frequently asked questions
Is 50 references enough for a 10,000-word dissertation? It can be, especially for an empirical or practice-based dissertation where your own data carries much of the argument. Fifty well-chosen, genuinely read sources that support every chapter will score better than ninety decorative ones. Check the spread and the quality of the set rather than the total alone, and check your brief for any minimum requirement.
Is 100 references too many for a 10,000-word dissertation? Not necessarily. A literature-heavy dissertation, or one spanning two fields, can legitimately reach one hundred references. It becomes too many when sources are cited without being discussed, when several sources repeat the same point, or when your own voice disappears under the weight of citation. If every reference does distinct work, one hundred is fine.
How many references should the literature review chapter have? For a literature review of around three thousand words inside a 10,000-word dissertation, roughly thirty to fifty references is typical. The exact figure depends on whether you discuss a smaller number of studies in depth or map a broader field more briefly. Depth of engagement matters more than the count.
Do references in the reference list count towards the word count? Usually the reference list itself is excluded from the word count at UK universities, while in-text citations are normally counted as part of the main text. Practice varies by institution, so check your own handbook before you rely on either rule.
Should I add references just to reach a higher number? No. Adding sources you have not read, or that add nothing to your argument, weakens the dissertation and creates an academic integrity risk if an examiner probes them. Improve the argument first; the reference count will follow the reading the argument actually needs.
Does my supervisor’s advice override general reference guidelines? Yes. Your brief and your supervisor know your department’s expectations and your topic. General ranges are for planning and sanity-checking; specific guidance about your dissertation always takes priority.