Getting permission to reuse a published figure or block of text is a licensing question before it’s a paperwork question. The right first step depends on how the original was published — a formal request through a rights-management tool is often unnecessary, and sometimes the request form itself is the wrong tool for the job. This guide walks through the decision in order: check the license, work out whether you actually need permission at all, use the publisher’s permission workflow if you do, and attribute correctly either way.
CASRAI’s Publication permission dictionary entry defines the underlying concept — the circumstances under which copyright material requires the copyright holder’s consent before reuse. This guide covers the practical process of getting there.
Start with the license, not a request form
The first question is never “how do I ask permission” — it’s “does this content already carry a license that answers the question.” Two pieces of scholarly-communication content that look identical on the page can have completely different reuse rules depending on how the publisher licensed them.
Scholarly communication draws a real distinction here, one that predates the popularization of Creative Commons: gratis open access (free to read, but still fully copyrighted — reuse beyond a narrow legal exception still requires permission) versus libre open access (free to read and licensed for reuse beyond that, typically via a specific Creative Commons license). The distinction was formalized by Peter Suber and Stevan Harnad, two of the original drafters of the Budapest Open Access Initiative definition. In practice: an article being “open access” tells you it’s free to read. It does not, by itself, tell you whether you can reuse a figure from it — you have to check which kind of open access it is.
If the article, chapter, or dataset is under a Creative Commons license, read the specific variant before assuming a green light:
- CC0 / CC-BY generally permit reuse, including in a new publication, with attribution (CC-BY) or without any condition at all (CC0).
- CC-BY-NC blocks reuse in a context a court or the licensor would consider primarily commercial — worth pausing on for anything going into an industry-sponsored article, textbook, or paid publication.
- CC-BY-ND (“no derivatives”) is the one that most often trips people up for figure reuse specifically: reprinting a figure unmodified is usually fine, but cropping, recoloring, relabeling, or otherwise altering it can cross into “derivative” territory the license doesn’t cover.
CASRAI’s Creative Commons Licenses for Research Data guide covers the full CC license family, version considerations, and how funders and repositories treat each one in more depth — the same license logic applies whether the reused item is a dataset, a figure, or a block of text. See also the dictionary entries for Licence, Open access, and Reuse license.
If a libre-OA license genuinely covers your intended use, you don’t need to request permission at all — compliance with the license terms (usually attribution, per the section below) is the entire obligation. Skipping straight to a publisher’s permissions desk for CC-BY content is unnecessary friction in both directions.
For traditionally copyrighted content: the RightsLink / Copyright Clearance Center workflow
If the content is gratis-OA or fully subscription/copyrighted (no reuse license attached), most major scholarly publishers route permission requests through RightsLink, a tool operated by the Copyright Clearance Center (CCC) and licensed by Elsevier, Wiley, Springer Nature, the American Chemical Society, AIP Publishing, SAGE, and others. The mechanism is consistent across publishers even though the entry point’s label varies:
- Locate the specific article, chapter, or book on the publisher’s own site (not a repository mirror or aggregator).
- Look for a “Get rights and content,” “Reprints and Permissions,” or “Request permissions” link on the article page — on ScienceDirect, for example, it sits under the title and author list, next to the DOI.
- Create a RightsLink account and select the specific reuse: what you’re reusing (figure/table, full-text extract, whole article), the new work it’s going into, and the distribution (print, electronic, worldwide, etc.).
- Submit the request. Many standard academic reuses (a single figure into a thesis or another journal article, for instance) generate an instant license and confirmation by email at no charge; others carry a fee or require manual review.
Not every publisher uses RightsLink — some, including AAAS/Science, direct certain requests to an in-house permissions department instead. If there’s no rights link on the article page, check the journal’s author-services or permissions page directly rather than assuming reuse is blocked.
A related but narrower mechanism worth knowing about: the STM Permissions Guidelines, a reciprocal agreement among STM Association signatory publishers that lets one signatory reuse a small amount of another signatory’s material — up to three figures/tables per article and limited word counts — in a new work, without a separate permission request or fee, provided the new work is itself substantially original (the guidelines specify at least 70% newly commissioned content). This is a publisher-to-publisher framework, not a general reuse right for any context (it doesn’t cover reuse in a standalone thesis, presentation, or personal website), but it’s why some in-house permissions teams can clear a small figure reuse instantly without you ever seeing a RightsLink form.
When you actually need publisher permission — and when you don’t
Permission isn’t always required, even for fully copyrighted content. The main exceptions:
- It’s your own prior work. Authors can generally reuse their own previously published figures or text in a new context (a thesis, a review article, a book chapter) without asking again — but check the specific copyright transfer agreement or publishing agreement you signed, since some retain narrower author rights than others. See Copyright transfer agreement and Publishing agreement.
- It’s genuinely in the public domain — copyright has expired, or (in the US) it’s a work of the federal government.
- A fair use or fair dealing exception applies — but these are legal defenses assessed case-by-case, not a checklist you can satisfy in advance, and they vary meaningfully by jurisdiction.
On that last point specifically, the two doctrines a CASRAI reader is most likely to encounter work differently:
US fair use (17 U.S.C. § 107) is an open-ended, four-factor balancing test: the purpose and character of the use (nonprofit/educational and “transformative” uses weigh in your favor), the nature of the original work, the amount and substantiality of what’s taken (both quantity and whether you took the “heart” of the work), and the effect on the market for the original. There’s no bright-line percentage or word count that guarantees fair use — courts weigh all four factors together, per the U.S. Copyright Office’s own guidance.
Fair dealing, the equivalent doctrine in the UK and much of the Commonwealth, works differently: it’s purpose-limited rather than an open balancing test. Under the Copyright, Designs and Patents Act 1988 (notably s.29 and s.30), permitted purposes are limited to non-commercial research and private study, criticism/review/quotation, and reporting current events — reuse for a purpose outside that list isn’t covered by fair dealing at all, regardless of how little you take.
In practice, both doctrines are a stronger defense for a short text extract with proper citation than for reproducing an entire figure or table wholesale — figures and tables are often treated as complete, self-contained works rather than a small “amount” of a larger one. For anything going into a formal publication (rather than internal use, a classroom, or a conference talk covered by an institution’s own exceptions), the safer default is still to request permission through the publisher rather than rely on a fair use/fair dealing judgment call.
Attributing reused material once permission is granted
Attribution is required regardless of which path above got you to “yes” — the difference is what specific credit line you owe.
For CC-licensed material, Creative Commons’ own best-practice model is TASL — Title, Author, Source (a link to where the original lives), and License (which variant, linked to the license deed). A minimal compliant example: “[Figure title]” by [Author] is licensed under CC BY 4.0, with the title and author linking back to the original source.
For content reused under a formal publisher permission, use the exact credit line the permission grant specifies — publishers are often strict about wording. Elsevier’s standard required format, for example, is: “Reprinted from [Publication title], Vol/edition number, Author(s), Title of article/chapter, Pages, Copyright (Year), with permission from Elsevier.” Other publishers specify their own wording in the RightsLink license confirmation; follow it exactly rather than improvising a citation-style credit, since the credit line is itself a condition of the permission, not just good scholarly practice.
For fair use/fair dealing reuse, a standard academic citation to the original source is the baseline expectation, even though it isn’t a formal “permission” credit line.
A quick decision path
- Check what license (if any) the original was published under — don’t assume “open access” alone answers this.
- If it’s under a CC license (or similar libre-OA license) whose terms your intended use satisfies, comply with the license’s attribution requirement. No permission request needed.
- If it’s gratis-OA or traditionally copyrighted, check whether it’s your own prior work, public domain, or already cleared under a framework like the STM Permissions Guidelines.
- If none of those apply, and your use isn’t clearly covered by a narrow fair use/fair dealing exception, request permission — via RightsLink/CCC if the publisher uses it, or the publisher’s own permissions desk if not.
- Attribute using the license’s required format (e.g., TASL for CC) or the exact credit line specified in the permission grant.
Frequently asked questions
Do I need permission to reuse a figure from an open-access article?
It depends on which kind of open access. Gratis OA (free to read, no reuse license) still requires permission or a fair use/fair dealing justification. Libre OA under a compatible CC license (most commonly CC-BY) generally does not — compliance with the license’s attribution term is the obligation, not a separate request.
What is RightsLink, and is there always a fee?
RightsLink is a permissions-management tool operated by the Copyright Clearance Center and used by most major scholarly publishers. Not every request carries a fee — many standard academic reuses (a figure into a thesis, for example) are cleared instantly at no charge, while others involve a fee or manual review depending on the publisher and the specific reuse.
Can I reuse my own previously published figure without asking permission?
Usually yes, but check the copyright transfer agreement or publishing agreement you signed — the rights you retained as an author vary by publisher and by agreement, and some require you to still credit the original publication.
Do fair use or fair dealing let me reproduce an entire figure without permission?
It’s the riskiest case for either doctrine, since a figure or table is usually treated as a complete work rather than a small excerpt of one, and using the whole thing weighs against a fair use finding under the “amount and substantiality” factor. For formal publication, requesting permission is the safer default.
How long does a permission request take?
Requests placed through RightsLink are frequently resolved instantly with an automated license and email confirmation. More complex or non-standard requests handled outside the automated system can take considerably longer — some publishers advise allowing several weeks for those cases.
What happens if I don’t use the exact credit line a publisher specified?
The credit line is a condition of the permission grant, not just a citation nicety — using different wording can put you outside the terms you were actually granted. If a required format was specified, use it exactly.
For the broader standards and definitions this process sits inside, see the Scholarly Publishing cluster, and the dictionary entries for Copyright and Publication permission.
