I wonder whether there is any way to create a DOI for a paper that was uploaded to arXiv but not published somewhere else, and have the DOI points to the arXiv URL (e.g., not pointing to some researchgate page). While there are reasons that arXiv does not provide DOIs (Why does arxiv.org not assign DOIs?), I prefer to use the same identifiers for all my research papers.

  • 1
    The faq page is doi.org/faq.html and it is the first question there.
    – Dirk
    Commented May 31, 2016 at 11:44
  • 4
    Something being in an FAQ somewhere doesn't mean it's off-topic. For one, the FAQ does not open for me at the moment, so it's not even a universal source. Voted to reopen.
    – yo'
    Commented May 31, 2016 at 23:13
  • 2
    @Dirk: I would consider the answer from the FAQ incomplete - using RAs is not the only way of generating and using DOIs. Some open science repositories, such as Zenodo and figshare, provide these services for free and directly (that is, without need to interact with the RAs). Commented Jun 1, 2016 at 0:12
  • 1
    @AleksandrBlekh - no, the FAQ is correct. Ultimately, figshare etc issue dois via a RA - though as this is back-end infrastructure, it's pretty invisible to the end user. They're not operating outside that system. Commented Jun 1, 2016 at 7:36
  • 1
    @Andrew note the difference between "incomplete" and "incorrect". The FAQ is correct, but if you say A (only RAs issue dois from IDF's point of view), you have to say B (there are third parties that do this for you).
    – yo'
    Commented Jun 1, 2016 at 8:34

1 Answer 1


Ultimately, no, you can't. There's no infrastructure to say "I want to get a single DOI for XYZ arbitrary url".

DOIs for most scholarly publications are issued through CrossRef. CrossRef do not assign DOIs directly, but delegate this to members or their agents. Members, who are usually publishers, pay a fee to the central consortium, and agree to issue DOIs based on a fixed set of rules.

As you can see from these rules, the general sense of membership is "organisations issuing DOIs for stuff they control". While there isn't an explicit prohibition against assigning DOIs to third-party material (which surprised me!), point 4 comes pretty close:

  1. Members have an obligation to maintain the metadata and URLs associated with all registered DOIs.

There is also a general prohibition of 'duplicative' works including preprints:

  1. Crossref only registers DOIs for Definitive Works (or Versions of Record, if not formally published) but not for Duplicative Works, as defined in the Crossref Glossary. This means that only original scholarly material, for which there is no actual DOI at the time of submission, and no expected duplication in future, is admissible for Crossref DOI registration. Crossref does not permit multiple DOIs to be assigned to certain closely related versions of a work, and hence does not support assignment of DOIs to Pre-prints or Post-prints of Definitive Works or to the Personal Version or a Self-archived Copy of a Definitive Work. For the same reasons, materials for which DOI duplication can be reasonably anticipated, such as an Authors Original Draft of a work being prepared for publication, are not admissible for Crossref DOI registration.

Putting those together, it seems likely that CrossRef's terms would prevent a member agreeing to issue a DOI for a (potentially duplicative) work hosted somewhere out of their control.

So, could you get a DOI from someone else, outside of CrossRef? Probably not. DOIs are only issued through a number of central registration authorities (eg CrossRef, DataCite). These have fairly well-defined areas of activity (eg DataCite won't issue DOIs for publications) and, to the best of my knowledge, none offer a "DOI for an arbitrary URL" service.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .