How long should a research paper introduction be?
About 10 percent of the paper, which is where five university guides land independently. For a 2,000 word course paper that is one or two paragraphs, roughly 150 to 250 words. ANU gives a thesis the same share, so a 15,000 word masters introduction runs 1,500 to 2,000. Length is the easy part, though. What decides whether an introduction works is the gap, the sentence naming what the existing work has not settled.
Ten percent, and five guides that say so
Ask five university skills guides and you get one answer. The University of Hull says an introduction "should be approximately 10% of your word count" (Essay writing, introductions). Newcastle University says "Your introduction is about 10% of the total word count." Derby labels the section "Introduction (approx. 10% of the word count)" and hands the conclusion the same share. The University of West London spells out what that leaves: "Unless you have been given a specific word count for your introduction or conclusion, each section should only be 10% of your word count (20% in total). The remaining 80% of the word count should be for your main body." And Victoria University states the whole budget in a line: "The general rule is 10% of your wordcount for the introduction, 80% for your paragraphs, or body, and the last 10% for your conclusion."
The share holds at every scale, which is what makes it useful. The Australian National University applies it to a thesis: "The introduction to the whole thesis can make up roughly 10 per cent of the total word count", so an 80,000 to 100,000 word PhD carries "a 8,000 - 10,000 word introduction" and a 15,000 to 20,000 word masters thesis one of "1,500 - 2,000 words" (ANU Academic Skills, introduction length).
| Whole paper | Introduction at 10% | What that is in practice |
|---|---|---|
| 1,000 words | about 100 | One paragraph, and the gap is one of its sentences |
| 2,000 words | 150 to 250 | One or two paragraphs, no roadmap needed |
| 5,000 words | about 500 | Two or three paragraphs, a roadmap sentence at the end |
| 15,000 to 20,000, masters thesis | 1,500 to 2,000 | A section, sometimes with its own subheadings |
| 80,000 to 100,000, PhD | 8,000 to 10,000 | A chapter of its own |
ANU names the one common exception, and it is worth knowing before you copy a number out of the table. Some theses run a much shorter introduction, "eight to ten pages (1.5 spacing) or about 4000 words for a PhD", in order to give a broad overview of the project before a long background or literature review chapter does the heavy work. If a separate literature review chapter follows yours, half the share is normal.
Length matters far less than structure. A tight introduction that moves cleanly from context to gap to thesis beats a long one that meanders through everything you know. If yours is sprawling past a tenth of the paper, the usual cause is that you have started writing the literature review inside it. Pull that material out and let the introduction stay lean, because its job is to set up the paper, not to be the paper.
What an introduction is actually for
An introduction has one job: convince the reader that your paper is worth their attention and tell them precisely what it will argue. It answers three questions in order. What is this about and why does it matter? What do we already know? And what is this paper going to add? A reader who finishes your introduction should be able to state your topic, your specific question, and your position before they read a single body paragraph.
That last part is the test. If someone reads your introduction and still does not know what you are claiming, the introduction has failed, no matter how well it is written. An introduction is not suspense. You give away the ending on purpose, because a research paper is an argument, and the reader needs to know the claim to follow the evidence for it.
The funnel: broad context to gap to thesis
Picture an upside-down triangle. You start wide, with the general subject, and narrow steadily until you reach the single point of your paper. This funnel shape is the backbone of almost every research introduction, and the reason it works is that it carries the reader from territory they already understand into the specific corner you are about to defend.
Academics formalize this as the CARS model, short for Create A Research Space, set out by the linguist John Swales in Genre Analysis (1990) after he studied how published research introductions are actually built. It breaks the funnel into three moves. First, establish the territory: show that the area matters and summarize what is known. Second, establish the niche: point out a gap, a tension, or an unanswered question in that known work. Third, occupy the niche: state how your paper fills the gap, which is your question and thesis. You do not need to memorize the jargon. You need the sequence: here is the field, here is what is missing, here is what I add.
The hinge of the whole thing is the second move, the gap. Most weak introductions skip it. They describe the topic, summarize some background, and then jump straight to a thesis with no reason the paper needed to exist. The gap is that reason. A single sentence, often starting with "however," "yet," or "but few studies," that names what the existing work has not settled. Without it, your thesis is an opinion floating in space. With it, your thesis is the answer to a problem the reader can now see.
The parts, in order
The funnel breaks into five moves. Not every introduction uses all five as separate sentences, and a short one may fold several together, but this is the order they come in.
Hook and context
Open with the broad subject and why it matters. This is the widest part of the funnel. A statistic, a real-world stakes statement, or a plain claim about the field's importance all work. Keep it honest and relevant; a dramatic hook that has nothing to do with your actual question is worse than a plain one that does. The goal is orientation, not theatrics. The reader should finish the first two sentences knowing what general area they are in and why it is worth their time.
Background: what is known
Narrow to the specific research conversation your paper joins. Summarize, briefly, what prior work has established, with citations. This is not a literature review; it is two or three sentences of the most relevant findings that set up your gap. Resist the urge to dump everything you read here. You are building a ledge for the reader to stand on, just high enough to see the gap you are about to point at.
The gap or problem
Name what the existing work has not done, cannot explain, or leaves contested. This is the pivot of the introduction and usually the hardest sentence to write, because it requires you to actually know the literature well enough to see its edge. Signal it clearly: "However, few studies have examined," "Yet it remains unclear whether," "This research has largely overlooked." The gap is the engine. Everything before it sets it up; everything after it follows from it.
Research question and thesis
State the specific question your paper answers, then your thesis, the arguable claim that is your answer. This is the narrow point of the funnel, the single idea the whole paper supports. The thesis should follow directly from the gap, so the reader sees that you are filling exactly the hole you just identified. If the question itself still feels broad, sharpen it first with the narrowing step in how to write a research paper, then come back. If you are unsure whether your thesis is sharp enough, we go deep on that in how to write a thesis statement.
Roadmap (optional)
In longer papers, close with one sentence previewing the structure: "This paper first reviews the street-by-street evidence on tree canopy, then presents the temperature records, before weighing alternative explanations." A roadmap helps the reader in a thesis or long report; in a short course paper it can feel mechanical and you can usually drop it. Use it when the paper is long enough that the reader needs a map, and the roadmap is easiest to write once you have an outline in front of you, since it just names the sections the outline already fixed.
Write it after the body
The single most useful habit is to draft the introduction last, or nearly last. An introduction promises what the paper delivers, and you cannot make that promise honestly until the paper exists. The UNC Chapel Hill Writing Center gives the same advice, on the reasoning that your argument often shifts as you write, so an introduction drafted first previews a claim you may not end up making. Write it first and you end up previewing an argument you had not made yet, then either rewriting the introduction or, worse, bending the paper to fit it.
A working introduction before you draft is fine as a target, a placeholder thesis and a rough gap to aim at. But the real introduction comes after the body, when you know what you actually argued, which evidence held up, and where the paper landed. At that point the gap, the thesis, and the roadmap can match the paper precisely, because you are describing something that already exists. This is the same reason the abstract is written last, which we cover in how to write an abstract.
A worked example
Here is the funnel built one move at a time, for a short paper on street trees and urban heat. Watch each sentence narrow the focus, and notice that the gap is the sentence the thesis answers.
Context (hook): "A dense city can sit several degrees warmer than the countryside around it on a summer afternoon, and the gap is widest after sunset, when paving and brick give back the heat they stored all day."
Background (what is known): "A large body of research finds lower air and surface temperatures on streets with heavy tree cover than on nearby streets with almost none, and several studies report that the gap is widest during the afternoon peak."
The gap: "However, almost all of this work compares different streets at a single moment, leaving it unclear whether canopy cools a street or whether the low-rise, low-traffic streets that stay cooler anyway are simply the ones with room for trees."
Question and thesis: "This paper asks whether adding canopy lowers the temperature of the same street over time. It argues that blocks that gained tree cover over the past decade cooled relative to blocks that did not, even when building density is held constant, because shade keeps heat out of the paving that would otherwise release it after dark."
Roadmap: "The paper first reviews the street-by-street evidence, then presents the decade of temperature records, before weighing building density and traffic volume as alternative explanations."
Read those five sentences in a row and you have a complete introduction. It starts wide, narrows through known research, pivots on a clear gap, lands on an arguable thesis that answers exactly that gap, and points the way forward. The hook here states the stakes instead of quoting a figure, which is the other way to open; in the finished paper it and every claim in the background move would carry a citation the reader can open. The shape is the part you copy; the content is yours.
Citations belong in the introduction too
An introduction without citations reads as opinion. The background move stands on what prior research has shown, and the gap is only believable if the reader can see the work it sits against. So the context and background sentences carry references, exactly like the body, and the same standard applies: every one has to be real.
This matters most in the introduction, because it is where a fabricated source hides best. A made-up reference dressed as established background looks completely at home next to a claim like "several studies report." If you are using AI to help draft, this is the riskiest spot for an invented citation to slip in unnoticed, and it is the reason AI makes up citations in ordinary chatbots: they generate plausible references from memory rather than retrieving real ones. Confirm each source exists before you lean on it, and never let an unchecked reference anchor your gap.
Common mistakes
Most weak introductions fail in one of a few familiar ways:
- Starting too broad. "Since the dawn of time, humans have wondered..." wastes the reader's first impression. Open near your actual topic, not at the origin of the universe.
- No gap. Describing the topic and then stating a thesis with no problem in between leaves the reader wondering why the paper exists. Name what is missing.
- No thesis. An introduction that ends without a clear, arguable claim sends the reader into the body with no idea what you are proving.
- Citing nothing. Background claims with no references read as assertion. The introduction needs sources like every other section.
- Writing the whole literature review. The introduction sets up the paper; it is not the place to summarize every source. Keep the background to what the gap needs.
- Writing it first and never revising it. An introduction drafted before the body promises a paper you had not written. Write it after, or rewrite it to match.
Draft the funnel, keep the sources real
Writing the introduction is still your thinking. Naming the gap and staking the thesis are judgement calls a tool should not make for you. Where CiteOwl helps is getting from your outline and sources to a first draft you then sharpen, on a verify-first principle: it searches actual literature and reads what it finds, so the background it summarizes and the references it places are real, not generated from memory. That is the difference that matters in an introduction, where a fabricated source hides most easily.
Give it your question and the sources you have gathered, and it drafts an introduction that follows the funnel, with each citation pointing at a paper it actually read and every change arriving as a reviewable diff you accept or reject. You keep the gap, the thesis, and the final words. If you want the larger picture of how the introduction fits the whole build, we lay it out in how to write a research paper.