How long does it take to write a literature review? Time yours

9 October 2026 · 7 min read

There is no honest single answer, because a literature review for a course paper and one for a thesis chapter are different jobs. What you can get is your own answer: break the review into stages, time the first batch of sources, and multiply by the size of your list. Much of the time goes into finding, reading and sorting sources before a sentence of the review exists, which is why estimates based on word count tend to come up short.

Why a literature review is hard to estimate

Most writing can be estimated from length. A literature review cannot, for three reasons.

• You do not know how many sources there are until you search. A tight question might turn up thirty relevant papers. A broad one might turn up three hundred, and you will not know which until you have screened them.
• Reading depth varies. Some sources need an abstract and a skim of the results. Others need two careful reads and a page of notes.
• Synthesis is the slow part, and it is invisible. Working out how a dozen studies relate to each other is thinking time, and it does not look like progress on the page.

The kind of review matters as well. A grant section, a dissertation chapter, a standalone review article and a systematic review with a written protocol differ widely in effort. Before estimating, write down which kind you are writing and roughly how many sources your supervisor, journal or instructor expects. If you do not know, ask. It is the biggest single input into the estimate.

The stages, and what each looks like on a Mac

Every literature review goes through the same stages, though the proportions differ. Each one tends to happen in a recognizable place on your Mac:

1. Searching. Running queries in library databases and search engines. Mostly the browser.
2. Screening. Reading titles and abstracts to decide what stays. The browser or your reference manager.
3. Reading. Working through the full texts. A PDF reader such as Preview.
4. Note-taking and synthesis. Summaries, a comparison table, themes. A notes app or a spreadsheet.
5. Writing. The draft itself. Your word processor.
6. Revising and citations. Restructuring, filling gaps, fixing references. Back and forth between all of the above.

The point of naming the stages is that each one has its own rate, and you can measure them separately.

Measure your own pace on the first batch

You do not need to time the whole review to estimate it. Time the first batch, then scale up.

1. Finish the search first. Run your searches and get a list of candidates. Note how long that took; you will repeat some of it later as you follow references.
2. Screen the whole list. Screening is fast per source, so time it across the full list and divide to get minutes per abstract.
3. Read and take notes on ten sources. Choose ten that are typical, not the easiest. Record the time spent reading and the time spent on notes.
4. Divide. Reading and note time divided by ten gives you a per-source figure.
5. Multiply by your list. That gives the reading stage for the whole review.
6. Estimate the writing from your own past work. If you have written a chapter or a long paper before, you know roughly how many hours a thousand words takes you when the research is done. Use that figure, then add half again for revision and citations.

Here is how the arithmetic works, with made-up numbers. Suppose reading ten papers took six hours and notes on them took two more. That is about forty-eight minutes per source. If sixty sources survive screening, the reading stage is about forty-eight hours. That is far more useful than “a few weeks,” because it shows where the calendar actually goes.

For a shorter assignment, How long does it take to write a 10-page paper? uses the same approach on a smaller scale.

Letting the Mac keep the count

The weak point of any self-timing method is the timing itself. You start a timer for a reading session, get pulled into an email, forget to stop it, and by the second week the records are mostly guesses.

Punchcard does the counting for you. It notices which app is in front through the day and, at the closing time you set, prints a receipt: the top five apps with time on each, the rest grouped as MISC, and a day total. For a review, that might read as Preview for the reading, the browser for searching, your word processor for drafting and a spreadsheet for the comparison table. Those lines map neatly onto the stages above.

If you search in Chrome, Brave, Edge or Vivaldi, you can switch on “Itemize websites” in Settings and the browser line splits into the sites you used, so your library database and journal sites each get their own line, competing with apps for the top five. It keeps only the site name, never the full address, search terms or page title. It does not work in Safari, Firefox or Arc; there the browser stays one line. The CSV export has one row per session with a start, an end and the app, which is enough to total each stage over a whole week in a spreadsheet.

What it cannot do: it does not know which paper you were reading, and it sees nothing you do away from the Mac. If you read printed articles or annotate on a tablet, add that time by hand. Work that happens away from the Mac, and why the record misses it has a simple way to do that.

Common reasons it runs long

When a review takes far longer than planned, it is usually one of these:

• The question is too broad. If screening keeps turning up relevant papers, narrow the question before reading more.
• Summaries instead of synthesis. Writing a paragraph about each source, in order, produces an annotated bibliography. Grouping sources by theme first makes the writing faster and the review better.
• Re-reading. Notes that are too thin send you back to the PDF. Spend an extra few minutes per source the first time.
• Citations left to the end. Fixing references across a long draft is slow. Add them as you write.

If your review is part of a larger degree, How much of a PhD week is actually writing? looks at where the rest of the week goes.

Questions

How many sources should a literature review have?

It depends entirely on the level and the field, so take the number from your instructor, supervisor or the journal’s recent reviews rather than from a general rule. Once you have it, it plugs straight into the estimate above.

Can I write a literature review in a week?

Time ten sources and do the multiplication. If the reading stage alone comes to more hours than you have, the answer is no, and you know it on day two rather than day six.

Should I count time spent thinking away from the desk?

Yes, if you want a true total. Synthesis often happens on a walk. Note it at the end of the day, roughly; a rough number is better than leaving it out.

Does the estimate get better over time?

It does. After one full review you have real rates for every stage, and the next estimate starts from those instead of a guess.