How long a review article should be depends first on which of the four review types you are writing, and second on the specific journal or publisher you are submitting to — there is no single industry-wide word count. This page covers the length angle specifically. For how to actually plan, structure, and write the manuscript itself, see CASRAI’s guide to how to write a review article, and for help deciding which review type your project actually is, see the comparison of literature review vs. systematic review vs. scoping review.
Why there is no single answer
Two variables drive review-article length, and they are independent of each other:
- Review type. A narrative review, a systematic review, a scoping review, and a meta-analysis carry different reporting obligations — a systematic review’s methods section alone (search strategy, screening, risk-of-bias assessment) typically runs longer than an entire narrative review’s discussion of the literature, simply because PRISMA 2020 and similar reporting standards require documenting steps a narrative review has no obligation to describe.
- Journal or publisher policy. Word limits (where they exist at all) are set by individual journals and publishers, not by a shared cross-industry standard. The same review type can have a stated limit of under 3,000 words at one journal and no fixed limit at all at another.
Because of this, “how long should my review be” only has a useful answer once you know both the type and the target journal. The ranges below are a starting orientation, not a substitute for checking the actual, current author instructions of the journal you intend to submit to — author guidelines are revised between submission cycles, and submitting against a stale figure is an avoidable cause of desk rejection.
Typical word-count ranges by review type
These are commonly observed ranges for the main text (introduction/background through discussion or conclusion), before journal-specific limits are applied on top:
| Review type | Typical main-text range | What drives the length |
|---|---|---|
| Narrative review | Roughly 3,000–6,000 words | No mandated methods-reporting section; length is set mostly by topic breadth and the target journal’s general article-length norms, since a narrative review does not follow a registered protocol. |
| Systematic review | Roughly 3,500–8,000 words | PRISMA 2020 reporting items (search strategy, screening flow, risk-of-bias assessment, results per outcome) add substantial required content beyond a narrative synthesis, even before a formal meta-analysis is included. |
| Scoping review | Roughly 4,000–8,000 words | Follows the PRISMA extension for scoping reviews (PRISMA-ScR); tends to run comparably to or longer than a systematic review because scoping reviews map a broader question and often present more descriptive/charting detail rather than a narrower, outcome-focused synthesis. |
| Meta-analysis | Adds to, rather than replaces, systematic-review length | A meta-analysis is a statistical synthesis layered onto (usually) a systematic review’s search and screening process — see CASRAI’s systematic review vs. meta-analysis comparison. Expect the combined systematic-review-plus-meta-analysis paper to sit at or above the upper end of the systematic-review range, driven by additional statistical-methods and results reporting (effect models, heterogeneity statistics, forest plots). |
These ranges describe common practice, not a formal standard set by any single body — treat them as a planning baseline, then confirm against the specific journal’s current instructions before drafting to a target length.
How ranges vary by journal and publisher
A handful of publishers state a specific, checkable word limit; most do not, and instead ask authors to keep length “appropriate to the content” or to confirm scope with an editor before submission. Examples of each pattern:
- Cochrane sets an explicit, numeric limit: as of July 1, 2026, new Cochrane systematic review and update submissions are capped at 12,000 words, and new protocol submissions at 6,000 words. For reviews and updates, the word count includes the abstract, plain-language summary, background, objectives, methods, results, discussion, and author conclusions — a wider scope than most journals count toward their limit. Cochrane’s review-authoring software (RevMan) now warns authors at 10,000 words for reviews/updates and 5,000 words for protocols, ahead of the hard cap.
- JAMA distinguishes Systematic Reviews from Narrative Reviews in its own author instructions and applies a materially shorter main-text limit to the systematic-review category than Cochrane’s own cap — commonly cited at around 3,500 words for the systematic-review category, well under half of Cochrane’s ceiling, reflecting JAMA’s format as a shorter clinical-journal article rather than a full evidence-synthesis publication. Confirm the current figure directly against JAMA’s live Instructions for Authors before drafting to it, since medical-journal limits are revised between cycles.
- BMJ and PLOS journals generally do not publish a single fixed word-count ceiling for review articles across the board; length expectations are set per article type and are typically described in qualitative terms (appropriate to the content, comparable to other accepted papers of that type) rather than a hard number. Always check the specific journal’s (not just the publisher’s) current author instructions, since policy can vary by title even within one publisher.
- Nature Reviews journals generally ask authors to discuss target length with the handling editor rather than publishing one fixed figure; general Nature review-format guidance has cited a main text around 5,000 words with a limited number of display items and a citation list capped well below what a Cochrane-style evidence synthesis would carry.
- Annual Reviews titles (the Annual Review of … series) are structured as long-form, invited review articles and routinely run substantially longer than the clinical-journal figures above — commonly in the tens of thousands of words across dozens of print pages, reflecting a comprehensive-field-survey format rather than a single-question synthesis. Consult the specific Annual Review title’s author handbook for its current length expectation rather than assuming one Annual Reviews figure applies across the whole series.
The pattern across all of these: journals that require a registered, replicable search-and-screening process (Cochrane, and increasingly PRISMA-aligned medical journals) tend to define length precisely, because the reporting burden itself is precisely defined. Journals publishing narrative or broader-scope reviews more often leave length to editorial judgment, because there is no equivalent fixed reporting checklist driving a minimum content requirement.
What usually counts toward a stated limit — and what doesn’t
Where a journal does publish a numeric limit, check exactly what it applies to before you draft toward it — this varies enough between publishers that assuming one convention across all of them is a common source of over- or under-length submissions:
- Main text (introduction/background through discussion or conclusion) is almost always inside the limit.
- Abstract is sometimes inside the limit (Cochrane explicitly includes it) and sometimes counted and capped separately with its own, smaller word limit (common at clinical journals, which often cap abstracts around 250–300 words independent of the main-text limit).
- References, tables, figures, and figure legends are usually excluded from a stated word-count limit, though many journals cap the number of references, tables, and figures separately — a reference-count limit is a distinct constraint from a word-count limit and both can apply at once.
- Supplementary material (extended search strategies, additional tables, a PRISMA checklist) is typically outside the main word limit entirely, which is why authors writing to a tight cap often move detailed methodological documentation there rather than cutting it. See CASRAI’s guide to the PRISMA 2020 reporting checklist for what that documentation needs to cover regardless of where it’s placed in the final submission.
Practical guidance for hitting a target length
- Confirm the number before you draft, not after. Pull the current author instructions for your specific target journal (not the publisher’s general guidance, and not a secondary aggregator site) and note the exact figure, what it includes, and whether abstract/references are counted separately.
- Match your review type’s real reporting obligations to the target. If you are running a full PRISMA-reported systematic review inside a 3,000–3,500-word clinical-journal limit, plan from the start to push detailed search strategies, full risk-of-bias tables, and extended results into supplementary material rather than discovering the shortfall during revision.
- Don’t pad a narrative review to look more substantial. A narrative review’s value is the quality of synthesis and argument, not length; reviewers and editors notice padding, and CASRAI’s own guide to writing a review article covers how to structure a narrative review’s argument so it earns its length rather than filling it.
- If you’re combining a systematic review with a meta-analysis, budget length separately for the two parts — the systematic review’s search-and-screening reporting and the meta-analysis’s statistical-methods-and-results reporting are both individually substantial, and combined papers are a common place where authors underestimate total length until a full draft is assembled.
- Recheck the limit at final submission, not just at the outline stage. Journal word limits are revised between submission cycles (Cochrane’s own July 2026 change is a direct example), so a figure that was correct when you started drafting months earlier may no longer be current by the time you submit.
Frequently asked questions
Is there one universal word-count standard for review articles?
No. Word-count limits for review articles are set individually by journals and publishers, not by a single cross-industry standard. Reporting standards like PRISMA 2020 specify what a systematic review must document, but they do not themselves impose a word count — the numeric limit, where one exists, comes from the journal.
How long is a systematic review compared to a narrative review?
A systematic review is typically longer than a narrative review on the same general topic, mainly because of the required methods reporting (search strategy, screening flow, risk-of-bias assessment) that a narrative review has no equivalent obligation to include. See the ranges in the table above, and CASRAI’s literature review vs. systematic review vs. scoping review comparison for how the two differ beyond length.
Does a meta-analysis have its own separate word limit?
Not typically as a standalone category — a meta-analysis is usually published as part of a systematic review paper, so it falls under that paper’s overall limit rather than carrying a separate figure of its own. See CASRAI’s systematic review vs. meta-analysis comparison for how the two relate.
What happens if my manuscript is over a journal’s stated word limit?
Practice varies by journal: some systems reject a submission automatically at the technical-check stage if it exceeds the stated limit, others flag it for the editor to assess case by case. Cochrane’s current policy states manuscripts exceeding its 12,000/6,000-word limits cannot be submitted at all from July 1, 2026 onward. Treat a stated limit as a hard constraint to design toward from the outline stage, not a target to negotiate after a full draft is written.
Related CASRAI resources
- How to Write a Review Article — full guide to planning, structuring, and writing a narrative review for submission.
- Literature Review vs. Systematic Review vs. Scoping Review — how to tell which type your project actually is.
- Systematic Review vs. Meta-Analysis — how the two relate and differ.
- PRISMA 2020 and Systematic Review Methodology — the reporting checklist behind a systematic review’s structure and length.
- Systematic Review Abstract: A Worked Example — abstract-specific length and structure guidance.
- Living systematic review — a continuously updated review format with its own length and maintenance considerations.
