Convert RIS references to BibTeX
The RIS format is the export format used by reference managers such as
EndNote, Zotero and Mendeley, and by databases like PubMed and Scopus. BibTeX
is the bibliography format LaTeX uses. This tool reads RIS records and produces
ready-to-paste .bib entries so you can move a literature library into a LaTeX
document without retyping anything.
The most common scenario: you have exported dozens of papers from PubMed or a
journal database as a .ris file, and you need them in a references.bib for
a manuscript or thesis. Doing it by hand means reformatting every author, year,
title and journal name. This converter does it in one paste.
How it works
RIS is a line-based format. Each line is a two-letter tag, two spaces, a dash, a
space, then the value — for example TI - The title. A record starts at a TY
line (the reference type) and ends at an ER - line. The parser splits the input
on those boundaries, collects each tag’s value (authors AU/A1 accumulate into a
list), then maps tags onto BibTeX fields: TI→title, PY/Y1→year, JO/JF/T2→journal,
VL→volume, IS→number, DO→doi, and SP/EP→pages joined with --.
The BibTeX entry type comes from TY: JOUR→@article, BOOK→@book,
CONF→@inproceedings, THES→@phdthesis, everything else→@misc. The cite key
is the first author’s surname plus the year (Smith2021); duplicate keys get a
trailing letter so every key is unique.
Example
A minimal journal record like this:
TY - JOUR
AU - Smith, Jane
TI - On parsing references
JO - Journal of Tools
PY - 2021
VL - 12
SP - 3
EP - 17
ER -
becomes:
@article{Smith2021,
author = {Smith, Jane},
title = {On parsing references},
journal = {Journal of Tools},
year = {2021},
volume = {12},
pages = {3--17}
}
Handling multiple authors
When a record has several AU lines, each author is included and they are joined
with BibTeX’s and separator — for example Smith, Jane and Jones, Robert. RIS
already stores names in Last, First order, which BibTeX accepts directly.
Common issues and how to fix them
Cite key collisions. If two papers share the same first author and year, the
converter appends a, b, c automatically. You can rename keys afterwards in
your .bib file; just update every \cite{} call in your document too.
Missing journal name. RIS databases sometimes omit JO and use T2 or JF
instead. The parser tries all three; if none is present the journal field is blank.
Fill it in manually or ask your database to export with JO populated.
Unknown reference types. Any TY value not in the recognised set falls back to
@misc, so no record is lost. You may want to change these to a more specific
entry type once you see the output.
Abstract field. The AB tag in RIS becomes the abstract field in BibTeX.
Many LaTeX styles ignore it by default, but tools like BibLaTeX can display it.
Workflow tips
- Export from PubMed using the PubMed format option, then switch the download type to RIS (it is labelled “Citation Manager” in some views).
- Zotero and Mendeley can also export selected items as RIS. Select the references you need, right-click and choose Export.
- After conversion, paste the
.bibcontent into your existingreferences.bibfile. BibTeX tolerates duplicate keys but will warn; use the cite-key collision suffix to resolve any clashes before you run LaTeX. - The conversion runs entirely in your browser — no RIS data is uploaded or stored.
Related tools
If you are already working with BibTeX and need to go in the other direction, the
BibTeX to CSL JSON converter lets you move references
into formats used by Pandoc and other citation engines. The
BibTeX Cleaner tidies up existing .bib files by removing
duplicate keys, normalising author formatting and stripping fields that LaTeX styles
ignore.