How to choose keywords for a research paper: Steps, examples, tips

Keywords are one of the smallest sections of a manuscript and one of the most consequential. They decide whether the right readers ever find your work, whether databases index it correctly, and, in many editorial systems, which reviewers are matched to it. Yet most authors assemble them in a hurry at the end of the submission process. This guide explains what keywords are, why they matter, how databases and search engines use them, and a clear, repeatable method for choosing, testing, formatting, and placing them so your paper reaches the audience it deserves.

Glossary of Key Terms

A quick reference for the terms used throughout this guide.

What Are Keywords in a Research Paper?

Keywords are the essential terms or concepts that embody the main topics, ideas, and scope of a study. In a published article they usually appear directly below the abstract, and you also enter them into the journal’s online submission system. A keyword such as “circular economy,” for instance, immediately signals that a paper concerns sustainability and waste reduction.

Their job is twofold. First, they tell readers and editors what the paper is about at a glance. Second, they act as tags that help search engines and academic databases identify your manuscript and surface it to people searching for related topics. In that sense they function a little like social-media hashtags for scholarship: compact labels that connect your work to a wider conversation. Crucially, keywords are not the same as a list of the words you happen to use most; they are chosen deliberately for retrieval.

Why Keywords Matter

With millions of articles published every year, even excellent research can be buried. Keywords are a primary mechanism by which your paper rises out of that pile. They matter for several concrete reasons:

• Databases and search engines scan titles, abstracts, and keywords to match articles to a query. Missing the terms your readers actually use means your paper may never appear in their results.
• Discoverability and citations are linked. Work that is easier to find is read more, and read more often leads to being cited more. Well-chosen keywords are associated with higher visibility and citation potential.
• Correct categorization. Keywords help databases file your paper under the right subject, so it reaches a relevant rather than random audience.
• Reviewer and editor matching. Many submission systems use your keywords to assign your manuscript to an appropriate editor and to locate qualified peer reviewers. Inaccurate keywords can route your paper to the wrong desk.
• For databases that sort by relevance, strategic keywords can be the difference between the first page of results and obscurity.

Treating keywords as just a box to tick at submission does your paper a disservice. A few minutes of deliberation can materially change how many people encounter your work.

How Databases and Search Engines Use Keywords

Databases are organized collections of scholarly articles drawn from many journals and publishers. Well-known examples include Scopus, Web of Science, and Google Scholar. When an article enters a database, it is indexed so that others can find it. That indexing draws on two kinds of terms:

• Author-supplied keywords: the terms you provide below the abstract and in the submission form.
• Database-derived terms: terms pulled automatically through full-text indexing or from the titles of cited references.

Some search engines, such as Google Scholar, may scan an article in its entirety. But the underlying principle is constant: if the critical terms are absent, the article is far less likely to surface. Keywords act as filters that let a database sift thousands of records down to the most relevant matches for a query.

Controlled vocabularies and MeSH

Some databases, especially in medicine, rely on controlled vocabularies: standardized term lists that provide a consistent way to find articles on a topic no matter which synonyms or abbreviations an author used. The best-known is MeSH (Medical Subject Headings), maintained by the U.S. National Library of Medicine and used by PubMed and MEDLINE. The MeSH thesaurus is also a practical source of established, popular terms. Where a controlled vocabulary exists for your field, aligning your keywords with it improves retrievability and consistency.

How to Choose the Right Keywords: A Step-by-Step Method

The quality of your search results is directly tied to the quality of your keywords. The following repeatable workflow brings together the best practice from editors, publishers, and university research guides.

Step 1: Understand the core of your study

Identify the central theme, research question, and scope of your paper. Write down the terms that capture its essence: these are usually nouns. Focus on what is genuinely central, not concepts that are barely touched.

Step 2: Mine the title and abstract

Extract the most essential terms from your title and abstract, since these sections carry the densest signal of what your paper is about. As an example, a paper titled “Impact of Nanobiosensor-Derived Lactate Monitoring on Intrapartum Hypoxia Outcomes in High-Risk Pregnancies” yields strong candidate keywords such as intrapartum hypoxia, high-risk pregnancy, nanobiosensors, lactate monitoring, and fetal acidosis.

Step 3: Brainstorm synonyms and related terms

Think about the different words other researchers might use to search for your topic, then list similar, broader, narrower, and related terms. A thesaurus, a concept map, or even describing a relevant image can help. So can asking a supervisor, a colleague, or a librarian. Keep the list beside you and keep adding to it.

Examples of useful substitutions and expansions:

• Botanical or species names → common names (Zea mays → maize; Oryza sativa → rice).
• Chemical or process terms → synonyms (degradation of a pollutant by microorganisms → bioremediation; adrenaline → epinephrine).
• Broader or narrower scope (vertebrates → mammals or primates; “renewable energy” → “sustainable energy”).
• Gene symbol in the title → full gene name as a keyword, and vice versa.

Step 4: Build a candidate pool, then narrow it

Assemble a generous pool of terms and word clusters that recur in your manuscript, keeping your target audience in mind. Then filter aggressively. Drop very general terms like “diet,” “medicine,” “technology,” “treatment”. These just bury your paper among countless others. Make them specific instead: “cancer treatment” is better, “liver cancer treatment” better still. Be cautious with newly coined terms and unusual abbreviations that few people will search.

Step 5: Prefer specific phrases over single words

Single words are often too broad and produce false matches. Two- to four-word phrases are usually more precise and effective. For a paper on leg pain, for instance, alternatives like leg cramps, muscular leg pain, or leg spasms will stand out far better than the generic word “pain.” Avoid conjunctions such as “and” within a keyword.

Step 6: Decide whether to mirror the title

Many journals automatically index the words in your title, so repeating them as keywords is a wasted opportunity. Use that space for synonyms and adjacent terms instead, so you get indexed for a wider range of searches. Some journals state this explicitly. (A handful do want overlap: always defer to the journal’s guidance.)

Step 7: Handle abbreviations carefully

Use an abbreviation only if it is well known, has a single dominant meaning, and is something researchers actually search. “ANOVA,” “PTSD,” and “HbA1c” (for a diabetes paper) are generally safe. Ambiguous ones are risky: “SDI” can mean strategic defence initiative or selective dissemination of information; “ARC” has meanings across many fields; “CGMP” could be continuous glucose monitoring program or current good manufacturing practice. When the abbreviation already appears in the title, use the spelled-out form as the keyword.

Step 8: Consider naming your study design

If the method is a defining feature of the work, include it: for example, “randomized controlled trial” or “systematic review.” Researchers frequently filter by design, so this can attract exactly the right readers.

Step 9: Use tools and controlled vocabularies

Lean on resources that reveal real search behavior: the MeSH thesaurus for medical and health topics, keyword-planning tools such as Google Keyword Planner, and the indexes of major databases. Some university libraries also offer keyword generators. These tools surface established terminology and trends rather than leaving you to guess.

Step 10: Test your keywords

Before committing, paste each term individually and in combination into the databases your field uses, such as Google Scholar and PubMed. Scan the results. If most are relevant to your study, the keyword works; if the results are off-topic, revise it. A classic trap: do not use “nursing” as a synonym for “breastfeeding,” because results will be dominated by professional nursing care: use “human lactation” instead.

Quick reference: dos and don’ts

How Many Keywords Should You Use?

There is no single universal number, but the common range is roughly three to ten, with most journals asking for about five to eight. A practical rule of thumb when your target journal is undecided is to prepare four to five strong keywords and expand from there. Whatever the count, make the selection span your main topics, relevant subdisciplines, and key methods so you cover a broad-enough spread without padding. Most importantly, the journal’s own rule overrides every general guideline: some cap the number, and some require single words rather than phrases.

How and Where to Use Keywords in Your Paper

Choosing strong keywords is half the task; placing and deploying them well is the other half. Used naturally throughout the manuscript, they reinforce discoverability without harming readability.

Title and abstract

These are the first things readers and search engines see. Work your most important terms into the title and weave key terms into the abstract, since abstracts are widely accessible even when the full text sits behind a paywall. Keep the title concise and readable; balance searchable terms against clarity. (Remember the title-overlap rule above when you compile the separate keyword list.)

Across the main sections

Use your keywords and their variations naturally across the introduction, methods, results, discussion, and conclusion. Descriptive subheadings that contain key terms help readers scan and help search engines parse your structure. The watchword is naturalness: the text must still read well.

Variations and synonyms

Beyond your shortlisted list, use related terms and synonyms in the body so the paper is found by readers who phrase things differently. But make sure that the writing isn’t repetitive or spammy.

Learn from highly cited work

Look at how well-cited papers on similar topics use their keywords. If a search for your intended terms surfaces strongly relevant articles, you know the terms work; their choices may also suggest keywords you hadn’t considered.

Avoid stuffing and irrelevance

Resist the urge to cram in extra or loosely related terms. Irrelevant keywords mislead readers and can exclude your paper from the searches that matter, and overstuffing can be penalized as keyword spam. A few precise, relevant terms used strategically beat a long, padded list every time.

Using Keywords to Search the Literature

Keywords are not only for the papers you publish: they are also how you find others’ work during your own literature review. The same vocabulary you build for your paper powers efficient searching.

• Broad search: start with the single most important term to gather background and get the lay of the land.
• Specific search: combine concepts with the operator AND to zero in on evidence for a particular claim.
• Too many irrelevant hits? add more terms to narrow the results.
• Too few relevant hits? remove or swap terms, or try synonyms from your list.

Good places to harvest search terms include your initial research questions, the encyclopedia and review articles you read for background, and the reference lists at the ends of books and articles. A concept map can help you organize broader, narrower, and related terms visually.

Formatting Keywords to the Journal’s Style

Keyword formatting is one of the most neglected details, yet journals differ on nearly every point. Before you submit, check the journal’s Information for Authors page: or simply inspect a few recently published articles: and match their conventions exactly.

• Label: some journals write “Keywords,” others “Key words” (one word or two).
• Style of the heading: Keywords, KEYWORDS (all caps), bold or plain, italic or plain.
• Punctuation after the label: usually a colon or a space.
• Capitalization: every keyword capitalized, only the first capitalized, or all lowercase.
• Separators: commas, semicolons, bullets, or just spaces.
• Layout: some journals list each keyword on its own line; in your manuscript you can simply separate them with spaces and let the journal apply final formatting.

Common Mistakes to Avoid

• Leaving keywords until the last minute and treating them as a formality.
• Using single, broad words where specific phrases would perform better.
• Simply copying the title into the keyword field when those words are already indexed.
• Choosing ambiguous abbreviations that mean different things in different fields.
• Inventing self-styled terms instead of the established vocabulary readers and reviewer-matching systems expect.
• Overstuffing with too many or loosely related keywords.
• Never testing the terms in a real database before submitting.
• Ignoring the journal’s formatting and number requirements.

Key Takeaways

• Keywords are search and indexing tools, not a summary; they help databases and readers find, retrieve, and cite your work.
• Most journals ask for roughly three to eight keywords; always check the specific author guidelines first, as some impose limits or require single words.
• Favor specific two- to four-word phrases over single, broad words such as “diet” or “technology.”
• Do not simply repeat words already in your title: those are usually auto-indexed. Use synonyms and adjacent terms to widen your reach.
• Mine your title and abstract for the essential nouns, then brainstorm synonyms readers might actually type.
• Use abbreviations only when they are unambiguous and widely recognized in your field (e.g., DNA, HIV), never ones with multiple meanings (e.g., SDI, ARC).
• Align with controlled vocabularies such as MeSH where they exist, and use planning tools to gauge real search demand.
• Test every keyword in the databases your readers use; if the results aren’t relevant to your paper, revise the term.
• Format keywords exactly to the journal’s style: capitalization, punctuation, and separators differ between journals.
• Keywords also drive reviewer matching in many submission systems, so accuracy affects who evaluates your paper.

Frequently Asked Questions

Does anyone actually use the keyword field, or is it a relic?

It is a fair question, and authors debate it often. Even where human readers rarely click a keyword, the field still does real work behind the scenes: many databases use it for indexing and filtering, and a great deal of editorial software uses your keywords to match your manuscript to subject editors and to search for suitable peer reviewers. So even if you think no reader will ever browse by keyword, the terms can still influence who handles and evaluates your paper: reason enough to choose them with care.

Should I pick the most frequent terms in my paper, or unique terms to game the search algorithm?

Neither extreme works well. Picking only your most repeated terms tends to produce broad words that you share with thousands of competing papers. Trying to “hack” the algorithm with obscure, unique terms means nobody searches for them. The reliable middle path is to choose specific phrases that genuinely describe your study and that real researchers are likely to type. And then verify that assumption by testing the terms in a database.

Will my older or pre-2000 paper be disadvantaged by keyword-based indexing?

Some indexing and recommendation systems lean toward recent terminology, which can make older work or studies that predate today’s vocabulary harder to surface. You can’t change a database’s algorithm, but you can help yourself by using the terminology that is current in your field now, including established controlled-vocabulary terms, so your work connects to the language readers are actually searching with today.

My field doesn’t have a MeSH-style controlled vocabulary. What do I do?

Many disciplines outside medicine lack a formal thesaurus. In that case, build your own reference by studying the keywords and recurring terminology in recent, well-cited papers in your area, and by checking the subject categories your target journal uses. The goal is the same: align with the language your community already uses to search, rather than inventing terms.

Can I reuse the same keywords across several of my own papers?

You can reuse genuinely relevant terms, and some overlap is natural if you work in one area. But each paper should still have keywords tailored to its specific contribution. If every paper carries an identical generic set, they compete with one another and none stands out for its distinct findings. Tune the specific phrases like method, population, or sub-topic to each study.

Do keywords differ for conference papers, theses, or preprints?

The principles are the same wherever your work is indexed: capture the core topics with specific, searchable terms. Conference and preprint servers and thesis repositories are indexed by search engines too, so keywords still aid discovery. Always follow the specific submission system’s rules on number and format, which can differ from journal norms.

How is a keyword different from a title or an abstract?

The title names the paper, the abstract summarizes it, and keywords are retrieval tags optimized for how people search. Because titles are often auto-indexed, keywords are most valuable when they add synonyms and adjacent terms the title and abstract don’t already contain, widening the set of queries that can find your work.

What’s the single most important habit for getting keywords right?

Testing. Before you submit, run each term through the databases your audience uses and look at what comes back. If the results match the kind of paper you wrote, the keyword earns its place; if not, replace it. This one habit catches ambiguous abbreviations, overly broad words, and misleading synonyms in minutes.

This article was published on August 9, 2022, and updated on June 10, 2026.