Boolean search string generator

Guide written by the AI for Research editorial team · Last reviewed 14 November 2026

A Boolean search string joins the synonyms for each concept with OR, joins the concepts with AND, and puts each OR group in parentheses: (teacher* OR educator*) AND (burnout OR "emotional exhaustion"). OR widens the search, AND narrows it, and NOT removes records. The builder above writes the string for PubMed, Europe PMC and a generic form most databases accept, and checks it for unbalanced quotes, stray operators and other errors before you paste it.

How do Boolean operators work?

AND finds records that contain every concept, OR finds records that contain any of the terms, and NOT removes records that contain a term. Most databases read the operators only in capitals [2].

Parentheses matter because databases differ in how they read a mix of AND and OR. PubMed reads a search from left to right [2], so nurse OR midwife AND burnout is not the same as nurse OR (midwife AND burnout). Put every OR group in brackets and every database reads it the same way.

How do you build a search string step by step?

Split your question into two or three concepts, list the words authors use for each, join the words with OR and the concepts with AND. The Cochrane Handbook describes the same structure for review searches [1].

1. Pick the concepts. Usually the population and the intervention or exposure. The PICO question builder gives you a concept table and sends it here.
2. List synonyms for each concept. Include spellings (behavior, behaviour), abbreviations (SMS, text message), plural forms and older terms. Read the titles and keywords of two or three papers you already know are relevant.
3. Decide on truncation and phrases. Truncate where the word stem is distinctive (midwi*). Quote multi-word terms you want kept together.
4. Choose the fields. Title and abstract is the usual choice for free-text terms. Searching all fields finds more, with more noise.
5. Add subject headings where the database has them. In PubMed, add MeSH headings next to your free-text words.
6. Check the string, run it, and test it. Make sure the papers you already know turn up. If one is missing, find out which term it uses and add it.

Worked examples across fields (illustrative)

Health: PubMed with MeSH

Concepts: nurses, night shift work, burnout. With the title and abstract field and MeSH rows, the builder writes:

("Nurses"[MeSH Terms] OR nurse*[tiab] OR "nursing staff"[tiab] OR midwi*[tiab]) AND ("Shift Work Schedule"[MeSH Terms] OR "night shift*"[tiab] OR "shift work*"[tiab]) AND ("Burnout, Professional"[MeSH Terms] OR burnout[tiab] OR "emotional exhaustion"[tiab])

Education: generic form

Concepts: teachers, burnout, mentoring. Generic output, ready for an education database:

(teacher* OR educator* OR "school staff") AND (burnout OR "emotional exhaustion" OR "teacher stress") AND (mentor* OR coaching OR induction)

Engineering: generic form, title only

Concepts: digital twins, maintenance, bridges. With Title only, the generic form stays the same and you add the database's title field code after you paste; PubMed would get [ti] and Europe PMC TITLE:.

("digital twin*" OR "virtual replica*") AND (maintenance OR "condition monitoring") AND (bridge* OR "civil infrastructure")

Psychology: Europe PMC

Concepts: mindfulness apps, university students. Europe PMC output with the title and abstract field:

(TITLE_ABS:mindfulness OR TITLE_ABS:"meditation app*") AND (TITLE_ABS:student* OR TITLE_ABS:undergraduate*)

Syntax for PubMed, Europe PMC and other databases

The concepts and the logic stay the same in every database; the field tags, subject headings and some operators change.

In PubMed, a MeSH heading also searches the narrower headings under it, and typing a field tag, quotes or truncation switches off PubMed's automatic term mapping for that term [2]. That is why the builder adds MeSH headings as their own terms next to the free-text words. For subscription databases your library provides, start from the generic form and use that database's help pages for field codes; the search strategy guide shows a full translation.

Truncation, phrases and NOT: when to use them

Truncate distinctive stems, quote phrases you need intact, and avoid NOT unless you have tested what it removes.

• Truncation. child* finds child, children and childhood. Very short stems match unrelated words: cat* finds category and catalyst. Some databases also refuse or cap very short stems.
• Phrases. "night shift" keeps the words together. Without quotes, PubMed may map the words to other terms, and other databases search each word anywhere in the field. With phrases switched off, the builder writes (night[tiab] AND shift[tiab]), so both words must still be there.
• NOT. burnout NOT rats removes every record that mentions rats, including a human study that says "unlike studies in rats". Review search guidance treats NOT with caution for this reason [1]. If you use it, run the search with and without NOT and look at what was removed.

What does the syntax check look for?

It flags errors that make a database reject or misread the string, and warnings that are legal but often a mistake.

You can paste any string, including one you edited by hand, into Check any search string.

Common mistakes

• No brackets around synonyms. nurse OR midwife AND burnout is read differently by different databases. Write (nurse OR midwife) AND burnout.
• Too many concepts joined with AND. Each AND narrows the search. Searching outcomes and time frames often loses relevant studies [1]; screen for them instead.
• Too few synonyms. One term per concept misses papers that use another word. Collect words from known relevant papers.
• Copying from a word processor. Curly quotes and dashes break searches. Paste into a plain-text box first.
• One string for every database. Field tags and subject headings differ. Keep the concepts and translate the syntax.
• Not recording the search. Save the exact string, the database, the date and the number of results. The numbered lines view gives you a log you can report, as PRISMA-S asks [4], and a librarian can review it with the PRESS checklist [5].

How to search with AI for Research

The builder runs in your browser and costs nothing. To go further:

• Search these terms in AI for Research runs the main term of each concept as a keyword search in paper search, across 250M+ scholarly works, with filters for year, study type and open access. It needs a free account. Search, citations and the free tools cost nothing; signed-in AI features use credits, and the Free plan has 45 credits a month.
• Save the papers you find to your research library and compare them in the literature review matrix builder.
• New to PubMed? Read how to search PubMed.

Next steps

• PICO question builder: turn a question into search concepts
• Systematic review search strategy: MeSH, testing and translating a search
• How to make a PRISMA flow diagram
• Literature review matrix builder

Sources

1. Lefebvre, C., Glanville, J., Briscoe, S., et al. Chapter 4: Searching for and selecting studies. In J. P. T. Higgins et al. (Eds.), Cochrane Handbook for Systematic Reviews of Interventions. Cochrane. https://training.cochrane.org/handbook/current/chapter-04
2. National Library of Medicine. PubMed User Guide (Boolean operators, field tags, MeSH, truncation, automatic term mapping). https://pubmed.ncbi.nlm.nih.gov/help/
3. Europe PMC. Search syntax reference. https://europepmc.org/searchsyntax
4. Rethlefsen, M. L., Kirtley, S., Waffenschmidt, S., et al. (2021). PRISMA-S: an extension to the PRISMA Statement for Reporting Literature Searches in Systematic Reviews. Systematic Reviews, 10, 39. https://doi.org/10.1186/s13643-020-01542-z
5. McGowan, J., Sampson, M., Salzwedel, D. M., Cogo, E., Foerster, V., & Lefebvre, C. (2016). PRESS Peer Review of Electronic Search Strategies: 2015 guideline statement. Journal of Clinical Epidemiology, 75, 40–46. https://doi.org/10.1016/j.jclinepi.2016.01.021