I am citing a figure I have taken from a journal. In my figure legend do I only include the number citation or do I need to state the figure is taken from somewhere else?

i.e Figure shows a cat [1]


Figure shows a cat. Taken from [1]


  • Or maybe Figure [1] shows a cat.
    – GoodDeeds
    May 22 '20 at 14:15
  • @GoodDeeds good point!
    – Sam
    May 22 '20 at 14:31
  • 1
    Is this for a course assignment or something you intend to publish?
    – Bryan Krause
    May 22 '20 at 15:08
  • @BryanKrause This is for my PhD thesis
    – Sam
    May 22 '20 at 15:37
  • 2
    @Sam Ok; make sure you have the necessary rights and permissions to reuse the figure (only a citation is not necessarily sufficient). If it's your own figure, most journals (and every one I am familiar with) have a policy stating it's no problem to reuse in a thesis and no special permission is needed, but you should check anyways. If it's someone else's then it depends on licensing and also on copyright laws by jurisdiction.
    – Bryan Krause
    May 22 '20 at 15:58

When you plan to reuse a figure taken from a journal paper, you should first ask permission to the copyright owner which, in most cases, is the journal publisher. Every publisher has its own procedure to make such a request, but usually you will be at least asked to specify where you want to republish the figure (in this case, specify PhD thesis).

The publisher will typically grant permission to reuse for free and it will also specify how it wants its copyright to be stated. Here are a couple of real examples:

Reproduced with permission from [1] ©IOP Publishing. All rights reserved.

© 2018 IEEE. Reprinted, with permission, from [1].

If your citation style is non-numeric, substitute the appropriate citation style.

In my experience, many people frequently overthink this procedure, as though obtaining the permission to republish could involve going through an obstacle course race, and sometimes they plan to spend hours to redo a figure which is just a few clicks away.

  • Thank you, it didn't even occur to me about permissions! I'll get that sorted right away
    – Sam
    May 26 '20 at 12:03

I think the most appropriate method is to give the information in the caption of the figure.

Fig 1. An example cat figure from the paper "Interesting Cats" by Johnson et al. [1]



Contrary to what the accepted answer says, it is not always necessary to get permission from the copyright holder (or author) to reproduce a figure (or quote a block of text). Check your local legislation, but most countries have a provision similar to the American "fair use" whereby reproduction for the purposes of scholarly criticism is allowed without any need for copyright request.

Notice, however, that something is not scholarly criticism just because it appears in a PhD thesis. I cannot give any good legal advice but if you are not saying anything more than what the author said in the original source, or if the figure could be omitted without significantly impairing the ability to understand the text, you are probably in the red; on the other hand, if you point to parts of the data that were not analyzed in the original publication, or if you discuss the figure as representative of a previous state-of-the-art (and describe the known shortcomings), you are probably OK.

Furthermore, it is possible that some publishers will require permission even when not legally required. I doubt thesis committees take a close look at this, but journal editors might. I cannot tell either way since I have only ever published one paper with the inclusion of a preexisting figure (from someone of my research group and published in the same journal, so an exceedingly favorable case for an editor to overlook a transgression).


I use "(rest of the caption). Taken from [1]" when I reproduce a figure from a former publication without change, and "(rest of the caption). Adapted from [1]" when I add modifications to the figure (arrows pointing to interesting points of data, zooming on a part of the figure, rewriting physical symbols to match the conventions in context, etc.).

I cannot say how common that practice is, but it is what I used for my PhD thesis (which went through my advisors and committee), and I decided on it after imitation of other papers.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.