Some journals have their own bibliography styles (bst) for BibTeX.

If a journal does not have its own bibliography style, am I supposed to include the content of the bbl-file into the tex-file I am sending? Or can I send the bib-file instead?

Many journals offer LaTeX-templates showing the style of the article. (This is almost always true if a journal recommends using its own document class.) In some case I have seen that this template contains BibTeX style (.bst file). But in other cases BibTeX was not used in the template, instead it was shown what the preferred formatting of references looks like.

I know that if I were the person who has to do the final typesetting from the sources submitted by an author to conform the style of the journal, I would prefer to have the BibTeX-file. (It is easier to simply change the formatting of references by changing the choice of BibTeX style than changing each separate entry manually.)

If a journal does not explicitly mention the recommended BibTeX style, what would be preferable way to submit the paper? Should I send both, TeX source and BibTeX file, so that editor can change the references using BibTeX style of their choice, or should I simply send the TeX-file, where I include also the bibliography generated by BibTeX. (And, if necessary, I modify it manually to be in the style required by the journal.)

(So far I have done the latter, but I am not sure whether it was the correct choice.)


I'd send the .tex and .bib file, which seems by far the most logical thing to do, to both you and me at least. On explicit request I'll grudgingly provide the .bbl file.

This should make things easier for them, as you correctly suggest. It is easy and fast to convert a .bib to a .bbl, but the other way round is impossible.

If the journal has a workflow that can't handle .bib files, that seems their problem and not mine. They are the professional publishers who charge for the "added value" of the professional typesetting and printing service, right?

  • 2
    Professional typesetter/layout editor (for a diamond-access journal) here, and .bib are definitely easier to work with. The only exception is if your .bib is of the unpruned, thousand entries long, using only @misc kind, in which case I have to completely redo the bibliography and starting from the .bbl is the fastest way. Happened six times last year...
    – Luris
    Apr 14 '16 at 12:28
  • @Luris Thanks for your feedback from the journals' side of things! In the unfortunate case when you have to start from the .bbl, do you ask the author for it, or do you generate it yourself starting from the messy .bib file? Apr 14 '16 at 12:46
  • 1
    I used to start from the .bbl (so that I don't have to redo entries that won't be used), and search the title/author/year in a database like MathSciNet or zbMATH. They generally give a BibTeX version of the reference that I simply copy-paste in the new .bib, but it's not always of good quality (zbMATH in the 1970-1980 range is particularly bad) nor cheap (MathSciNet cost us 10k€/year). But now that we stopped our subscription to MathSciNet, I'll probably ask the author for a clean .bib, especially since they're asked to do it when they submit their paper :)
    – Luris
    Apr 14 '16 at 13:46

Most journals do not have the resources (read time) to handle lots of auxiliary files in submissions. Unless the journal clearly specifies they want bibliographies provided as a .bib-file and has a fixed .bst style file, you should not provide LaTeX based manuscript that way. What you can do is to run your final version yourself and then manually include the resulting .bbl bibliography in your manuscript so that content and bibliography is included in the same file (see e.g., How can I insert by .bib file into my .tex file? on TeX.sx). The .bbl file contains all references in the common \bibitem[]{} format. Alternatively you can of course use the \bibitem form directly when you write the paper. If you do not have a bibliography style file for the specific journal, it would be simple to use one that yields a similar format and then manually correct the details that may differ in the .bbl file.

  • 2
    Most journals? Where does that come from? Given the typical subscription cost, it would really surprise me if they "do not have the resources to handle lots of auxiliary files". Dec 13 '13 at 20:44
  • 3
    Transforming a .bib into a \bibitem or a .bbl takes minutes in worst cases. The other way can take hours, especially if the typesetter don't have access to a good database like MathSciNet.
    – Luris
    Apr 14 '16 at 11:52
  • @BenoîtKloeckner I know the ACM SIG Proceedings format requires you to do the conversion yourself before .tex submission, at least, but they make it explicit. (Admittedly, that's for conference proceedings rather than a journal, but still.)
    – JAB
    Dec 5 '17 at 14:46

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.