Examples
Worked examples
- Is an instance
A biomedical journal following the ICMJE Recommendations sends a submitted manuscript to two or three external reviewers with relevant expertise; the handling editor weighs their written assessments of the study's methods and conclusions against the author's response before deciding to accept, reject, or request revisions — the manuscript is not published until that cycle resolves.
- Is an instance
A Registered Report is peer reviewed twice: once at Stage 1, before data collection, when reviewers assess the study question and planned methodology, and again at Stage 2, after results are in, when reviewers check that the study followed the accepted protocol — a variant explicitly built to separate review of the idea from review of the outcome.
Counter-examples
Looks similar, but isn't
- Not an instance
A preprint posted to a server such as arXiv or bioRxiv has not been peer reviewed, even though it may carry a DOI and be citable — most preprint servers perform only a basic screening for scope and completeness, not independent expert evaluation of validity, and say so explicitly in their own policies.
- Not an instance
A magazine feature or trade-publication article may be edited and fact-checked by in-house staff, but without independent external subject-matter reviewers gating acceptance, that editorial review does not meet the definition of peer review.
Editorial commentary
Peer review is the process by which a journal or other scholarly venue submits a manuscript to independent experts in the relevant field for evaluation before deciding whether to publish it. The reviewers assess whether the work’s methods support its conclusions, whether the research is original and significant enough to warrant publication, and whether the manuscript meets the venue’s standards — and their written feedback, mediated by an editor, determines the outcome. A publication described as “peer reviewed” is one where this gatekeeping step is a matter of documented editorial policy, applied to the research articles it carries, not an occasional or optional practice.
Why it exists
Peer review exists to catch errors, gaps, and overstated claims that an author, closest to their own work, is poorly positioned to see — and to filter for significance and fit before a venue commits its editorial resources and reputation to a piece of work. It is not a guarantee of correctness (reviewers do not typically re-run experiments or re-analyze raw data, and disputes about a study’s conclusions regularly continue after publication), and the research-integrity literature is explicit that it is a quality filter, not a certification of truth. The ICMJE, for instance, frames peer review as valuable but not beyond dispute — a documented process, not an infallible one.
How the process works
Mechanics vary by publisher, but the common sequence is: an author submits a manuscript; an editor screens it for basic scope and quality and either desk-rejects it or sends it out for review; two or more independent reviewers with relevant expertise evaluate it and submit written reports, usually recommending accept, reject, or revise; the editor weighs those reports (and the author’s response to any requested revisions) and makes the final decision. Reviewers are typically unpaid volunteers drawn from the same research community as the authors, distinguishing journal peer review from paid grant-application peer review run by funders such as NIH’s Center for Scientific Review.
Identity-disclosure models
The most visible variation between venues is how much reviewer and author identity is disclosed to the other party during review — see CASRAI’s comparison of single-anonymized, double-anonymized, and open peer review for the differences between these models, and the individual entries for single-anonymized peer review, double-anonymized peer review, and equitable peer review. A shared vocabulary for describing these models and other peer-review metadata was formalized by NISO as ANSI/NISO Z39.106-2023, developed with STM (the International Association of Scientific, Technical and Medical Publishers).
Who does the reviewing, and what they owe the process
The person doing the evaluating is the peer reviewer — see CASRAI’s fuller guide on becoming and acting as a peer reviewer for the role, and COPE’s Ethical Guidelines for Peer Reviewers for the conduct expectations (confidentiality, disclosed conflicts of interest, timely and constructive feedback) that most editors hold reviewers to. The editor coordinating the process is the journal editor, typically supported by an editorial board of subject-matter experts who help select reviewers and adjudicate disagreements.
What counts as peer reviewed, and what does not
A finished, published journal article carrying a documented peer-review policy is peer reviewed. A preprint is not, even once it has a DOI and is citable — see CASRAI’s guides on how to know if an article is peer reviewed and how to find peer-reviewed journals for practical checks, and how to search specifically for peer-reviewed articles in a database search. Because “peer reviewed” is a policy claim a publisher makes about itself, readers sometimes need to verify it venue by venue — CASRAI covers this for specific large publishers, for example whether MDPI journals are peer reviewed and whether Taylor & Francis journals are peer reviewed. Review does not end at acceptance either: many venues now support post-publication peer review, and sites such as PubPeer host open post-publication commentary distinct from a journal’s own formal process. See also CASRAI’s broader explainer, Peer Review: What It Is, How It Works, and Why It Matters, for the fuller narrative version of this concept, and the entry on the academic journal as the publication type that peer review is a defining feature of.
Machine-readable encodings
Use in your systems
<role vocab="credit"
vocab-identifier="https://casrai.org/dictionary/"
vocab-term="Peer Review"
vocab-term-identifier="https://casrai.org/dictionary/term/peer-review" />{
"@context": "https://schema.org",
"@type": "DefinedTerm",
"@id": "https://casrai.org/dictionary/term/peer-review",
"name": "Peer Review",
"identifier": "https://casrai.org/dictionary/term/peer-review",
"description": "A work is peer reviewed when, before publication, an editor sends it to one or more independent experts in the same field (peer reviewers) who evaluate its methodology, validity, originality, and significance, and the editor uses that written feedback to decide whether to accept, reject, or require revisions. Three elements distinguish genuine peer review from ordinary editorial handling: (1) the reviewers are subject-matter experts external to the editorial staff, not merely copy-editors; (2) the review is documented (recorded reviewer reports, even if reviewer identity is withheld from the author); and (3) the outcome materially gates publication rather than serving as optional feedback offered after the fact. The specific procedure varies by venue — reviewer and author identities may or may not be disclosed to each other (single-anonymized, double-anonymized, or open review), and some venues review a study's protocol before results exist (registered reports) rather than only the finished manuscript — but the core requirement of independent expert gatekeeping prior to publication is common to all forms. A 'peer-reviewed journal' is simply a journal that applies this process, as a matter of documented policy, to the research articles it publishes.",
"inDefinedTermSet": "https://casrai.org/dictionary/domain/research-outputs#set",
"url": "https://casrai.org/dictionary/term/peer-review",
"sameAs": [],
"license": "https://creativecommons.org/licenses/by/4.0/",
"publisher": {
"@id": "https://casrai.org/#organization"
},
"dateModified": "2026-07-18T06:30:46",
"inLanguage": "en"
}
