BiomedicalPublications

How to verify a journal is legitimate

Seven checks, in the order worth doing them, that separate a real journal from one that will take your money and publish anything. Together they take about twenty minutes and require nothing but a browser.

We run the machine-checkable parts of this automatically against 11,447 biomedical journals. Run the check on a journal now — then use the steps below for the parts only a human can judge.

No single item on this list condemns a journal. Honest journals fail individual checks for boring reasons: a new title is not yet indexed, a small regional journal sits outside the major registries, a subscription journal has no DOAJ record because DOAJ only lists open-access titles. The signal is several failing at once.

The seven checks

  1. Find the ISSN, then verify it independently
  2. Check the indexes yourself, not the journal's claims
  3. Read the fee page before you submit, not after acceptance
  4. Look up three editorial board members
  5. Read a recent issue
  6. Be suspicious of speed
  7. Check what happens if the publisher disappears

1. Find the ISSN, then verify it independently

Every genuinely published journal has an International Standard Serial Number. It is the closest thing scholarly publishing has to a registration number, and unlike a logo or a badge it cannot be invented, because it is issued by a national ISSN centre and recorded in a public register.

Take the ISSN from the journal's own site — usually in the footer or on an “about” page — and look it up at the ISSN Portal. You are checking one thing: does the registered title match the title on the website you are looking at?

What failure looks like: no record at all; a record whose title is subtly different; or an ISSN that belongs to a different journal entirely, borrowed to look credible. A mismatched ISSN is the single most decisive red flag on this list, because there is no innocent explanation for it.

A print and an electronic ISSN are different numbers for the same journal, so two ISSNs is normal, not suspicious. Of the 11,447 journals in our directory, every one has at least one verified ISSN — it is our minimum condition for inclusion.

2. Check the indexes yourself, not the journal's claims

Journals routinely claim indexing they do not have. The claim costs nothing to make, sits on their homepage as a row of logos, and is almost never checked by the authors it is designed to reassure.

Search for the journal directly in the registries rather than trusting the badge: PubMed's journal catalogue for biomedical coverage, and DOAJ for open-access titles. Both are free and take under a minute.

What failure looks like: a homepage badge with no corresponding record; a claim of “PubMed indexed” that turns out to mean a handful of articles deposited in PMC by funders rather than the journal being indexed in MEDLINE; or invented index names that sound official but have no registry behind them.

Be careful with the distinction between being in PubMed and being indexed in MEDLINE. They are not the same, and the difference is exactly where misleading claims live. 3,620 of our journals have a full DOAJ record, which is where editorial metadata such as review model and licensing comes from.

3. Read the fee page before you submit, not after acceptance

This is where researchers actually get hurt. An article processing charge that appears only once the paper is accepted arrives at the worst possible moment: withdrawing means starting the submission process over, months later, and the journal knows it.

Before you submit, find the charge in writing, on the journal's own site, with a number attached. Not “fees may apply.” Not “contact us for details.” A figure.

What failure looks like: no fee page; a fee page with no number; a fee quoted only in email after submission; or charges that appear under a different name — submission fee, handling fee, page charge, colour figure charge — on top of the headline APC.

Across our directory 4,496 journals publish a charge, with a median of $3,140 and a maximum of $12,690. A further 1,290 state that they charge nothing at all. You can filter on this directly with the APC finder.

A journal that will not tell you the price before you commit has told you something about how it treats authors.

4. Look up three editorial board members

Pick three names from the editorial board — not the editor-in-chief, who is the one name most likely to be real, but three from the middle of the list. For each, ask three questions: does this person exist, do they work where the journal says they work, and do they publish in this field?

A search for the name plus their stated institution answers all three in about a minute. Their institutional profile page or ORCID record is the evidence you want.

What failure looks like: board members with no online footprint; affiliations that do not exist; researchers whose actual field is unrelated to the journal's scope; or a board listed only as names and countries, with no institutions at all.

There is a subtler failure worth knowing about: some journals list real, well-published academics who never agreed to serve and did not know they were listed. If a board member's own website or CV makes no mention of the editorship, that is worth a second look.

5. Read a recent issue

Open the most recent issue and read two or three articles properly. This is the step people skip, and it is the one that resolves most ambiguous cases faster than any registry lookup.

You are not assessing whether the science is correct. You are asking whether anyone edited this: are the articles on the journal's stated topic, is the English competent, are the figures legible, are the references real and relevant, is there evidence that a reviewer ever asked for a change?

What failure looks like: articles wildly outside the stated scope sitting side by side; obvious uncorrected language problems throughout; formatting that varies article to article because nothing was typeset; or a journal claiming a broad international readership whose entire recent output comes from one or two institutions.

A journal that publishes anything will demonstrate that to you in five minutes of reading. No amount of registry metadata substitutes for looking at the product.

6. Be suspicious of speed

Genuine peer review takes weeks, because it involves finding qualified people who are busy, waiting for them to read carefully, and giving the authors time to respond. There is no way to compress that to seventy-two hours while keeping the review.

Of the 3,620 journals that report an average time from submission to publication, the median is 14 weeks. Treat that as your reference point. An advertised turnaround dramatically below it is not efficiency; it is the absence of review.

What failure looks like: “acceptance within 72 hours” or similar as a selling point; guaranteed acceptance in any form; or an unsolicited email praising your recent paper and inviting a submission to a journal you have never heard of.

That last one deserves naming plainly: a flattering invitation email is a marketing funnel, not an honour. Legitimate journals do not recruit authors this way, and the flattery is generated from your public publication record.

7. Check what happens if the publisher disappears

The least intuitive check, and the one with the longest consequences. Ask two questions: are articles assigned DOIs, and are they deposited with a digital preservation service such as CLOCKSS, LOCKSS, or Portico?

A DOI is a permanent identifier that keeps resolving even if the article moves. Preservation means an independent archive holds a copy if the publisher shuts down. Both are standard practice, and both are usually stated on the journal's about or policies page.

What failure looks like: articles reachable only at a publisher URL with no DOI; no archiving statement anywhere; or a preservation claim that names no actual service.

Without both, your paper can simply vanish — and a vanished paper does not count for anything, in a CV, a grant application, or a promotion case. This is the check that matters in ten years rather than this week.

Putting it together

Work down the list and count the failures. One failure with an obvious explanation is usually nothing. Three or more — no verifiable ISSN, no registry record, no stated fee, no named board — is a pattern, and you do not need anyone to attach a label to it. You have your answer in a form you could defend to a supervisor or an ethics committee.

If a journal passes, you have also produced something useful: a documented reason for choosing it. Keep the notes with your submission record.

Why we do not publish a blacklist

You will find lists online naming journals and publishers as “predatory.” We do not maintain one. The best-known such list was taken offline in 2017 after sustained legal pressure from publishers it named, and lists of that kind age badly: journals change hands, editorial standards change, and a name once listed follows it indefinitely.

Verifiable facts are more useful and more durable than a label. That position is set out in full in our methodology and editorial disclosure.

Next steps