As a thirsty graduate student, I've amassed a burgeoning collection of research articles in both paper and digital formats. Right now I manage PDFs and references with Zotero, but I'm still searching for an organization logic for paper documents that allows rapid access and prevents redundant printing and storage. Some have suggested organizing by topic, while another approach is to sort by author name. Topical organization has not worked well for me in the past because my research is highly interdisciplinary, confounding my categorization efforts. For a coupled digital-physical organization system I'm considering the following:

  • Digital documents and references stored and tagged in Zotero
  • Indicator in Zotero whether or not I've printed in the file
  • Physical documents stored in manila file folders labelled alphabetically

What is the most effective way that you've found to maintain both paper and digital document repositories?


3 Answers 3


I have them organized similarly to the blog post you linked. I'm not sure it is the optimal way, but it works quite well.

  • The central search key for me is (First)AuthorYEAR. By now I know my way around in my field so that the author names are meaningful. And easier to remember exactly than title abbreviations.

  • I try to have almost all papers also digitally, and organize them with jabref. I started using jabref before Mendeley, Zotero & Co came up and didn't change.

  • jabref allows to assign the papers to topics/groups, which can be hierarchically organized, but a paper can also belong to several of them.
    This is important for me, because one paper may be about an application (e.g. tumor diagnostics -> group with tumour or cell type), use specific measurement technique (group that), and maybe use or invent an interesting data analysis technique (group there).

  • Jabref stores its information in a .bib file (with a few non-standard fields and some more info in comments at the end). But essentially I can work with this file as with any other .bib file.
    .bib and digital versions are in a git repo which basically solved the mess of having lots of copies that go out of sync while I'm still able to work completely offline. As .bib is a text format, version control works fine with that. Papers usually do not change, so it doesn't really matter for the version control that they are binary data.

  • When looking for a paper, I usually search for it in jabref, and then look whether I do have a printed copy (I'm too lazy to keep track of printed status).

  • physical copies I keep in two drawers with a suspension filing system. One has collections of important papers on topics I'm interested in. The other keeps other physical copies alphabetically by author name (one folder for each letter so far, some will be split soon).

  • I tried before with normal lever arch files, but was too lazy to put papers away into their proper place.

  • For a while I had the luxury of working with dual monitors. During that time I printed considerably less papers. Now I have to work with only a single (and not too big) monitor again, and again print most papers.


I always store the references in zotero before printing them. I don't print all of the references, therefore I always use zotero to find the references before searching for the printed paper.

In this context what I did is:

  1. When I print the paper I write in one corner the date and time when it was added to zotero (always the same corner).
  2. When I search for a paper I check that identifier (date-time) and perform binary search on the big pile of printed papers.

I only print papers that are very complex on formulas, though. This could not scale well for hundreds of papers, but it's a handy identifier. One advantage is that papers are usually added to one side of the pile, so old papers can be stored somewhere and I don't use to need to insert a new paper between old papers, if you use folders (or similar) some of them could overflow by doing that.


I always create a collection in Zotero for each of my projects and papers. That places a paper into a proper context. As was already mentioned above, the nice option of Zotero (and jabref) is that you may put a bibliographic record into infinite number of collections.

I hope, you know of and use zotfile and better bibtex extensions for Zotero. Zotfile makes sure the PDFs are stored and named properly; and better bibtex creates a .bib file for your library, just like jabref (that happens to be useful dealing with TeX).

You must log in to answer this question.

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