Journal Reviewer: Role Defined

Journal reviewers sit between authors and editors in the standard peer-review workflow. An editor decides a submission merits external, expert evaluation, invites one or more journal reviewers to assess it, and the reviewer’s formal output — a written report and a recommendation — goes back to that editor, not to the author directly and not into the published record. The reviewer does not decide whether the manuscript is published; that authority stays with the editor.

How reviewers are selected and invited

Per ICMJE’s “Responsibilities in the Submission and Peer-Review Process,” selecting appropriate reviewers — and making sure they can access everything relevant to evaluating the manuscript — is the editor’s responsibility, not the author’s and not a fully automated process. In practice, editors draw on several overlapping channels: their own knowledge of active researchers in the manuscript’s subfield, the journal’s own reviewer database, keyword/expertise matching built into the submission system (ScholarOne, Editorial Manager, and similar platforms), reviewers suggested by the submitting authors (subject to the editor’s independent vetting, precisely because an author-suggested reviewer carries an obvious conflict-of-interest risk), and self-nominated or volunteer reviewer pools some journals maintain. COPE’s Ethical Guidelines for Peer Reviewers frame taking up reviewing as a professional responsibility that follows from having benefited from being reviewed, not a single fixed application path — some journals require formal appointment to a review panel or a demonstrated area of expertise, and expect accurate personal and professional information from anyone accepting an invitation. A reviewer should only agree if they genuinely have the relevant expertise and can render an unbiased assessment, and should flag any gap in their expertise to the editor rather than review outside it.

Typical review criteria

What a journal reviewer is actually asked to judge varies by journal and by manuscript type, but most review criteria cluster around the same handful of questions: is the methodology sound and are the conclusions supported by the data (validity); has this been done before, and if so, is the contribution here genuinely new (originality); does the work matter enough to the field to justify publication (significance); and is the manuscript clearly written and adequately situated against prior literature (clarity). Many journals formalize these into a structured review form or rubric, and discipline-specific submissions add further checks — ethical approvals and consent statements for human- or animal-subject research, data and code availability, statistical reporting adequacy, and reproducibility of reported methods. Note that ANSI/NISO Z39.106-2023, Standard Terminology for Peer Review, does not standardize review criteria themselves; it standardizes how a journal describes its review model — identity transparency, who the reviewer interacts with, what review information is published, and whether post-publication commenting is enabled — a related but distinct piece of controlled vocabulary (see CASRAI’s ANSI/NISO Z39.106-2023 entry).

Conflict-of-interest and confidentiality expectations

Both COPE and ICMJE set out reviewer obligations in detail. COPE’s Ethical Guidelines for Peer Reviewers (Version 2, September 2017) require a reviewer to declare all competing interests — personal, financial, intellectual, professional, political, or religious — before accepting an invitation, and to decline if they are currently at the same institution as an author or have been a close collaborator, mentor, mentee, or joint grant holder with an author within roughly the past three years. On confidentiality, COPE requires reviewers to keep the manuscript and its contents strictly confidential throughout and after the review, never to use unpublished ideas or data from it for their own purposes before the work is published, and never to involve anyone else — including a trainee or mentee — in the review without the editor’s explicit permission. ICMJE’s parallel guidance is consistent: editors must not share information about a manuscript with anyone besides its own authors and reviewers, and reviewers “must keep manuscripts they are asked to review confidential” and disclose competing interests that could bias their assessment. Both bodies also flag a specific, named form of reviewer misconduct: suggesting that authors cite the reviewer’s own work without a genuine academic reason for doing so.

Signaling availability: how to get onto an editor’s reviewer list

Beyond waiting for an editor to happen to notice a relevant publication, there are three concrete ways a researcher can make themselves findable as a prospective journal reviewer:

A journal’s own reviewer database. Most journals and publishers maintain a self-nomination form (often under a “become a reviewer” or “reviewer resources” link on the journal’s website) where a researcher lists their subject-matter expertise areas directly; some submission systems (ScholarOne, Editorial Manager) also let a registered user flag reviewer availability and keywords inside their own account profile.
Web of Science Researcher Profile’s Peer Review Preferences. A researcher can use the Peer Review Preferences tab on their Web of Science Researcher Profile to indicate specific journals they are interested in reviewing for. Editors at partnered journals can see who has expressed interest and send an invitation directly through the platform; a researcher can also opt to share their email with a given journal so an editor can contact them, and can turn on automatic addition of verified reviews from partnered journals and funders to their profile.
An ORCID record with visible affiliations and works. Editors and editorial-system keyword matching draw on publicly visible researcher metadata to identify subject-matter experts; keeping an ORCID iD current with affiliations, works, and subject areas makes a researcher more findable through the same channels editors already use to locate reviewers, even without a dedicated reviewer-database entry.

Publishing corresponding-author papers in a subfield remains, by a wide margin, the most common route into that subfield’s reviewer pool — invitations from editors following a manuscript’s own reference list, or a recommendation from another reviewer or author, generate more first invitations than self-nomination does. The channels above are best understood as ways to be findable in addition to publishing, not a substitute for it.

Declining a review invitation well

Declining is a routine, expected part of the reviewer role, not a professional lapse — COPE’s guidelines are explicit that a reviewer should only agree to review when they have the necessary expertise and the time to meet the requested deadline, which means honestly declining is often the correct response to an invitation. Good practice when declining:

Respond promptly either way. Editors need to know quickly whether to look for another reviewer; a slow non-response is more disruptive to the editorial timeline than a fast decline.
State the reason briefly. A short, honest reason — outside my expertise, a competing interest with an author, unavailable in the requested window — is normal and expected; an invitation is a request, not an obligation, and editors do not need an elaborate justification.
Suggest an alternative reviewer if a genuinely suitable one comes to mind. Many submission systems include a field for this; it is a common, welcomed courtesy, not a requirement.
Decline outright — rather than accepting and passing the manuscript to someone else — if the underlying issue is a competing interest. COPE and ICMJE both treat involving another person in a review without the editor’s explicit permission as a confidentiality violation, so redirecting an invitation informally is not an acceptable substitute for a clean decline.

What a first review report should contain

A reviewer new to the role can reasonably structure a first review report around four parts, which map onto the review criteria most journals already use (see “Typical review criteria” above):

A brief summary of the manuscript in the reviewer’s own words — what the authors did and claim to have found. This is useful to the editor as a check that the reviewer and the authors understood the paper’s contribution the same way, and it is standard practice even though it is rarely explicitly requested.
Major comments — substantive issues that bear on whether the methodology supports the conclusions, whether the contribution is sufficiently original, and whether the significance justifies publication. COPE’s guidelines call for these to be objective and constructive, substantiated with specific evidence rather than general assertions (“the analysis is weak” is not actionable; pointing to the specific step that doesn’t follow is).
Minor comments — smaller issues: unclear wording, a mislabeled figure, an incomplete citation, a suggestion that would strengthen but is not required for publication.
A recommendation to the editor (commonly accept, minor revision, major revision, or reject, though the exact categories vary by journal) — remembering that this is advisory, not a decision; per ICMJE, “a peer-reviewed journal is under no obligation to follow reviewer recommendations.”

Most submission systems also provide a separate confidential-comments-to-editor field, distinct from the comments the authors will see. COPE’s guidelines are specific that this field should stay consistent with what is shown to the authors and should not become a place for accusations the authors never get to respond to.

Recognition for review work

Because journal review is typically unpaid and traditionally anonymous, recognition infrastructure is the primary way a reviewer gets credit for the work at all. CASRAI’s Peer Reviewer guide covers this in full, including how ORCID’s Peer Review record works (entries can only be pushed by a trusted organization via the ORCID API, not self-entered) and how the Web of Science Reviewer Recognition Service can auto-export verified review history to a linked ORCID record — both mechanisms depend on the journal or a recognition service acting on the reviewer’s behalf, so opting in and keeping an ORCID iD linked to the submission system is what actually makes review work visible after the fact.

How the reviewer role differs from an editor’s

A journal reviewer’s contribution is advisory, not decisive. ICMJE states this in unambiguous terms: “a peer-reviewed journal is under no obligation to follow reviewer recommendations” — the editor makes the final accept, revise, or reject decision and can diverge from what any individual reviewer, or even a majority of reviewers, advised. The editor also carries responsibilities no reviewer holds: selecting and inviting reviewers in the first place, resolving disagreement between reviewers who reach conflicting conclusions, managing the manuscript and correspondence throughout the submission’s life at the journal, and communicating the decision — with or without reviewer comments attached — back to the authors. A reviewer’s formal role begins and ends with the manuscripts they are specifically invited to assess; it is not a standing editorial position, and it carries no authorship credit or publication-decision authority in its own right.

For the full practical path into and through the role — how invitations actually reach a prospective reviewer, typical turnaround expectations, and reviewer-recognition infrastructure like ORCID’s Peer Review record — see CASRAI’s guide, Peer Reviewer: Role, Responsibilities, and How to Become One. For how the review stage fits into the broader editorial and submission process, see CASRAI’s What Is Peer Review? guide. For a side-by-side look at how identity-transparency choices (who knows whose name, and when) change what a review looks like in practice, see CASRAI’s single-blind vs double-blind vs open peer review comparison.

References

COPE Council, Ethical Guidelines for Peer Reviewers, Version 2 (September 2017)
ICMJE, “Responsibilities in the Submission and Peer-Review Process,” Roles and Responsibilities
NISO, ANSI/NISO Z39.106-2023, Standard Terminology for Peer Review (July 5, 2023)