Yes. It indexes copies that authors and institutions have deposited legitimately in repositories under their publisher agreements. It hosts nothing itself; it resolves a DOI to a legal free version where one exists.

Will an author actually send me their paper?

Usually. Most publisher agreements permit an author to share a copy for personal scholarly use, and a short, specific request gets a high response rate. Say who you are and what you are working on; a one-line request with no context does worse.

What is the difference between green and gold open access?

Gold means the publisher makes the version of record free, often funded by an article processing charge. Green means the author deposits a version — usually the accepted manuscript — in a repository, sometimes after an embargo. Green is what Unpaywall mostly finds.

Can I use a preprint if I cannot get the published version?

You can read it, and it is usually close to the published text, but check for changes made in review before relying on any specific number. Cite the version you actually read and label it as a preprint if that is what it was.

Getting papers you cannot access

Before paying thirty-nine dollars for a PDF, spend ninety seconds on the five checks below. They resolve most of the time.

How can I legally access a paywalled research paper?

Check Unpaywall or OpenAccessButton first, which find author-deposited copies automatically. Then the institutional or subject repository, Europe PMC and PubMed Central for funded biomedical work, the preprint version, and the author personal page. If none of those work, request it through interlibrary loan or email the corresponding author, who is usually permitted to send it.

Why a free version usually exists

Most publisher agreements let an author deposit some version of their own paper in a repository — often the accepted manuscript, which is the peer-reviewed text before the publisher typesetting. Many funders require it. The result is that a large share of paywalled papers have a legal free copy somewhere, and the difficulty is finding it rather than obtaining permission.

Version	What it is	Usually free?
Preprint	The manuscript before peer review	Yes, on a preprint server
Accepted manuscript	Peer-reviewed text, before publisher formatting	Often, in a repository, sometimes after an embargo
Version of record	The published, typeset article	Only if the article is open access

The accepted manuscript is the one to look for. Its content is the peer-reviewed content; only the layout, pagination and copyediting differ. Cite the version of record and note which version you read if the difference could matter.

The order that works

Unpaywall. A browser extension and an open API that checks over fifty thousand repositories from the DOI. It is the highest-yield single step and takes no effort at all.
OpenAccessButton and CORE. Different aggregation, so they sometimes find what Unpaywall does not.
Europe PMC and PubMed Central. For biomedical work, funder mandates put a very large share of it here, often after a twelve-month embargo.
The institutional repository. Search the author university repository directly, and BASE or OpenAIRE across many at once.
The preprint server. Check arXiv, bioRxiv, medRxiv, SSRN or the relevant server for the same title.
The author own pages. Personal sites, lab pages, ResearchGate and Academia.edu profiles — though the last two host material of uncertain legality, so prefer a repository copy where both exist.
Interlibrary loan. If you are affiliated with any institution, this is a normal service, usually free to you, and typically delivers within a few days.
Email the corresponding author. Say briefly who you are and why you want the paper. Authors are generally permitted to send a copy for personal scholarly use and are usually pleased to be asked. The response rate is high.

Reading a paper without the PDF

Sometimes you only need to know whether a paper is worth pursuing. Several things are available without access to the full text:

The abstract, from Crossref, OpenAlex, PubMed or Semantic Scholar.
The reference list, which OpenAlex and Semantic Scholar expose openly for a large share of papers.
The citation context — the sentences in later papers that cite it, which often state what it established more clearly than its own abstract does.
Any deposited dataset, protocol or supplementary material, which is frequently in an open repository even when the paper is not.
A related preprint or conference version by the same authors.

What not to do

Sites that host pirated publisher PDFs are infringing, and using them exposes institutional networks to blocking and, in some jurisdictions, individual users to liability. There is also a practical objection: the routes above resolve most requests within a day and do not require anyone to accept that risk.

If a paper is genuinely unobtainable and central to your work, record that. In a systematic review it is reported as a report not retrieved in the PRISMA diagram, which is a limitation to be stated rather than a gap to be quietly closed.

Questions

Is Unpaywall legal?: Yes. It indexes copies that authors and institutions have deposited legitimately in repositories under their publisher agreements. It hosts nothing itself; it resolves a DOI to a legal free version where one exists.
Will an author actually send me their paper?: Usually. Most publisher agreements permit an author to share a copy for personal scholarly use, and a short, specific request gets a high response rate. Say who you are and what you are working on; a one-line request with no context does worse.
What is the difference between green and gold open access?: Gold means the publisher makes the version of record free, often funded by an article processing charge. Green means the author deposits a version — usually the accepted manuscript — in a repository, sometimes after an embargo. Green is what Unpaywall mostly finds.
Can I use a preprint if I cannot get the published version?: You can read it, and it is usually close to the published text, but check for changes made in review before relying on any specific number. Cite the version you actually read and label it as a preprint if that is what it was.

Finding and using preprints — Where can I find preprints?
Google Scholar alternatives — What are the best alternatives to Google Scholar?
Searching grey literature — What is grey literature and how do you search it?
How to find research papers on a topic — How do I find research papers on a topic?

Last updated 2026-08-25. Part of the litscout literature search guides.