Journal Keyword Rules: How Many to Pick, and Which Ones.
Waraq · July 21, 2026 · RULE OF THE WEEK
Most journals ask for somewhere between three and eight, and the exact number is set by the journal you submit to, not by any shared convention. Frontiers in Neurology asks for five to eight. Read your target journal's author guidelines for its range before you draft the list, because the count is a submission requirement the form can reject you against.
Are keyword counts an actual rule, or just advice?
They are a rule. Many journals write a specific range into their author guidelines, and the online submission system enforces it.
The keyword field in a submission form usually carries a minimum and a maximum. Enter four when the journal wants five and the form may refuse to advance. Editorial staff also check the count during the first format screen, next to things like reference style and abstract length. A wrong count is the kind of small mismatch that can bounce a manuscript before a reviewer opens it, the same way an over-long abstract can. If you want to see how these mechanical limits stack up across journals, the abstract word-limit comparison walks through several of them.
What does Frontiers in Neurology require for keywords?
Five to eight, with a spread of broad and narrow terms. The rule, quoted from the Frontiers in Neurology author guidelines, reads: "Pick 5-8 keywords using a mix of generic and more specific terms on the article subject(s)."
Two instructions sit inside that one line. The first is the count. The second is the mix: some of your keywords should be wide, some should be tight. A paper on a rare seizure disorder might pair a general term such as epilepsy with the name of the specific syndrome and the gene involved. The wider terms bring in readers browsing the field; the narrower ones find the handful of people working on exactly your problem. The five-to-eight window leaves room for both. This is one journal's rule. Others set their own range, so treat the number as journal-specific rather than universal.
Should your keywords repeat words from the title?
Generally no. Many publishers advise against reusing title words, because databases already index the title, so a repeated word spends a keyword slot without adding a new route to your paper. Elsevier and Taylor & Francis both make this point in their author guidance, and some journals ban title-word keywords altogether.
Think of each keyword as a separate pointer to your article. Your title words are already pointers. Copying one into the keyword list wastes the slot. Spend it instead on a term the title had to leave out: the method you used, the population you studied, a synonym a searcher might type, or the broader field your work belongs to. A title that names a disease and an intervention still says nothing about your study design, so "cohort study" or "meta-analysis" earns its place as a keyword precisely because the title never mentions it.
How do MeSH and controlled vocabularies differ from free keywords?
MeSH, short for Medical Subject Headings, is a controlled vocabulary maintained by the United States National Library of Medicine and assigned by its own indexers, separate from the keywords you submit with the manuscript.
A controlled vocabulary is a fixed list of approved terms; a free keyword is any word you choose yourself. For biomedical articles, NLM staff read the full text and attach roughly a dozen MeSH headings to each MEDLINE citation, flagging a few of them with an asterisk as the major topics. You do not pick those, and they do not replace your author keywords. Knowing the controlled term for your subject still helps, because it tells you the standard wording searchers use, which you can then echo in the free keywords you do control. Other fields run their own schemes, such as the ACM Computing Classification System in computer science. When your discipline has one, borrowing its exact phrasing makes your paper easier to find than an idiosyncratic label would. The aims and scope of a journal often hints at the vocabulary its readers expect, which is another reason to read that section closely before you submit.
A submission-day keyword checklist
Run this pass once your draft is otherwise ready, in the order below.
- Open the author guidelines and note the required range. If none is stated, five keywords is a safe default for most journals.
- Count entries, not words. A two-word or three-word phrase is a single keyword.
- Strike any keyword that already appears in your title, unless the journal explicitly permits the overlap.
- Follow the mix instruction where there is one: pair a couple of broad terms with several specific ones.
- Match your wording to your field's controlled vocabulary when one exists, so your terms line up with how people actually search.
- Format as the guidelines say. Some journals want the list alphabetized, some want commas, some want semicolons.
Keyword count is one of a cluster of small, checkable rules that decide whether a manuscript clears the first screen or comes straight back. The same screen looks at reference formatting, spacing, and how many authors you list before the et al. cutoff. A last read against the journal's own format rules catches all of them together, and running the manuscript through Waraq surfaces the mismatches before an editor does. Fix the keyword list at that stage and it stops being a reason for rejection and starts doing its real job, which is bringing the right readers to your work.