How do I tell if a journal is predatory?

The old advice, watch for bad grammar and garish websites, has stopped working. Predatory publishing became a professional industry, and current sites are indistinguishable from legitimate ones on presentation. Some copy the branding of real journals closely enough to fool experienced researchers, including using a real journal's name with a slightly altered domain.

What still works is verification through independent sources. A legitimate journal is discoverable somewhere you can check: a curated index, a publisher that belongs to an ethics body, a listing in your library's databases. A journal that exists only on its own website and in the emails it sends you has not been verified by anyone.

Be precise about what you are accusing a journal of. Using the Committee on Publication Ethics definition, predatory publishing is the systematic, for-profit dissemination of purported scholarship through deceptive or fraudulent practices and without quality assurance (Leonard et al., 2021, p. 1). The test is the process, not the price or the business model. A survey of 1,007 top-quartile medical journals across 29 specialties found median article processing charges of $4,190 across 698 hybrid journals and $2,940 across 307 fully open-access journals (Tocco et al., 2025, p. 4), so a four-figure fee on its own tells you to investigate, not to walk away.

The unsolicited invitation is the strongest single signal. Real journals rarely email individual early-career researchers asking for submissions, and they never flatter you about a paper they have not read. An email praising your recent work and inviting a submission with a fast track is a form letter sent to thousands.

Treat blocklists as leads rather than verdicts. Beall's List was taken down in 2017, the versions that circulated afterwards were anonymous, and Cabells is not a complete solution on its own (Leonard et al., 2021, p. 4). A listing tells you to read the inclusion criteria and then check the journal's own practices, and absence from a list proves nothing, because new predatory journals keep appearing. When the evidence stays mixed, verify the publisher's physical address and contact details and check whether it belongs to COPE, ICMJE, OASPA, or C4DISC (Leonard et al., 2021, p. 4).

Ask a librarian. Research librarians deal with this weekly, they know the current patterns, and they will check a journal for you in about ten minutes. It is the fastest reliable answer available.

Which checks are actually reliable?

Four. Indexing in a curated database rather than a self-declared list. Publisher membership of a recognized ethics body. Editors who are real, contactable researchers in the field. And a peer review process with a plausible timeline, since a promise of review in 72 hours is a promise not to review.

The editor check is worth doing properly. Email one of the listed editorial board members and ask whether they are active on the board. Predatory journals routinely list researchers without their knowledge, and those researchers are usually glad to be told.

Check the article processing charge and when it is disclosed. A legitimate fee is published before submission. A fee that appears only after acceptance is a strong signal, and so is a fee that is unusually low for the claimed service.

The size of the fee decides nothing by itself. Legitimate journals may charge US$5,000 or more, and some open-access journals charge authors nothing (Leonard et al., 2021, p. 2). What a journal owes you before submission is a full disclosure of processing charges, page charges, translation fees, and any other cost, plus a clear answer on waivers. The APC survey above is narrower than it looks, since it covered only top-quartile Journal Citation Reports journals, excluded diamond open access and journals with $0 charges, and did not collect submission or review fees (Tocco et al., 2025, pp. 7-8), so do not treat its medians as the normal range for your field. If the journal advertises a Journal Impact Factor, look it up in Web of Science or Journal Citation Reports. A metric that is not in the database it names is a concrete finding, not an impression (Leonard et al., 2021, p. 3).

Judge the editorial board on evidence you can check. Named editors should carry institutional affiliations, their professional profiles should confirm they serve on the board, and their expertise should match the journal's stated scope. The same source warns you not to use institutional prestige, geography, or personal familiarity as a legitimacy test (Leonard et al., 2021, pp. 2-3), which matters because that reflex penalizes small independent journals and journals from under-resourced regions rather than the operations you are trying to catch.

What do I do if I already submitted?

Withdraw immediately and in writing, before acceptance if possible, and keep the correspondence. If they demand a withdrawal fee, that demand is itself confirmation. If the paper was already published, you generally cannot submit it elsewhere, and your institution's research office is the right place to go next.

Do not pay a withdrawal fee. There is no legitimate charge for withdrawing a manuscript before acceptance, and paying tends to produce further demands.

If the paper is out and you want it removed, ask formally in writing and keep records. Success is uneven, and your research office or library may have dealt with the same publisher before. This happens to careful people, and handling it openly is far better than hoping nobody notices.