About Dewey Lookup
Dewey Lookup is a free tool for the library of one — the school, church, community, prison, museum, or special library run by a single person (or a single person and a lot of goodwill). It answers one stubborn, daily question: what Dewey Decimal number does this book get?
If your library belongs to OCLC, you already have copy cataloging and you don’t need this. But OCLC membership is out of reach for a great many small libraries, and the free service OCLC once offered for exactly this purpose (Classify) was retired. That left a real gap. This tool fills the practical part of it.
Where the numbers come from
Dewey Lookup never makes a number up. It reads numbers that catalogers already assigned, from open bibliographic records. The tool works down a chain, stopping at the first catalog that has a Dewey number for the book in hand:
- Open Library (Internet Archive) — fast, free, and carries a Dewey number for a great many titles.
- HathiTrust — full catalog records contributed by more than fifty research libraries.
- Library of Congress — read straight from the catalog record’s 082 field.
- Harvard Library, then K10plus (the German union catalog) — often the ones that come through for international and UK editions.
- Three public-library consortiums — NExpress (Northeast Kansas), VOKAL (Vermont) and SEKnFind (Southeast Kansas) — asked together when nothing above has a number. This is copy cataloging from peer libraries that actually shelve the book.
Google Books doesn’t supply Dewey numbers, but it fills in the description, cover and subject headings that make the rest of the card useful.
Sibling editions can help too: hardcovers, paperbacks, and translations of the same work often carry different records, and one may have a Dewey number where another doesn’t. Edition matching powered by LibraryThing’s thingISBN service — a quiet gift to the library world that lets us check whether a book’s sibling editions carry the call number this one lacks.
A suggestion, when nothing is found
Everything above is copied. Someone at a cataloging library sat with the book, assigned a number, and this tool went and fetched it.
Sometimes no record exists at all — not for the book, not for any edition of it, in any of these sources. In that case the tool can offer a suggestion instead: a number worked out from the title and subject headings by software trained on a large body of already-cataloged books. The software is Annif, an open-source project from the National Library of Finland.
This is the one number on the page that no cataloger assigned. So it is labelled Suggested, shown with a confidence figure, and never dressed up as a found record. Treat it as a pointer to the right part of the schedules, not as copy cataloging. If a real record turns up for the book later, the real number wins.
A constructed number, for fiction
Fiction is the hard case. Novels are shelved as FIC by most libraries, so a great many fiction records carry no Dewey number at all — there is nothing to copy, even when the book sits in a dozen catalogs.
For fiction, an 800s literature number isn’t looked up — it’s constructed, the same way a cataloger builds one: from three facts about the author, not the story. Which national literature (the language the author writes in, and their country), which literary period (when the author flourished), and the form. Those facts come from the Library of Congress Name Authority File and Wikidata; the number is assembled from the published Dewey tables for literature.
Because no cataloger assigned it, a constructed number is labelled Constructed and shows which sources supplied the author’s facts. It is a defensible candidate to check against the schedules, not copy cataloging — and when the facts are uncertain (an author near a period boundary, or of two nationalities) it offers the alternatives and lets you decide. When the author’s facts can’t be established at all, it stays silent rather than guess. If a real record turns up later, the real number wins.
When there’s no record at all
Old, local, or esoteric items often have no record of their own. For those, the tool does what an experienced cataloger does by hand: you describe what the book is about in a subject or two, and it finds common, well-cataloged works on those subjects and shows you the Dewey numbers they were given — grouped by discipline — so you can borrow the number that fits the book in hand. It’s called classifying by analogy.
Two forms of the same number
You’ll often see the number twice: large at the top of the card, and again — longer — on the 082 line below.
The 082 line is the number exactly as the cataloging library recorded it, including any prime marks, like 323.119/6073/009046. Those slashes aren’t punctuation. They are the cataloger’s own guidance about where the number can safely be cut short, and they come from the Dewey schedules — you can’t work them out by looking at the digits. Not every library records them, but when one does, it’s telling you something useful.
The large number is a shelf-ready version of the same thing: cut to four decimal places. That cut is arbitrary. Four places is roughly the depth most small libraries shelve at — it isn’t a rule from the schedules, and it may fall in a slightly different place than a cataloger would choose. If the 082 line shows prime marks, they are the better guide to where to stop. Copy whichever form suits your shelf.
Who makes it

Dewey Lookup is built and maintained by David Thompson, a working school librarian and IFLA Personal Affiliate, who hit this exact wall — cataloging an English-language collection without OCLC — and got tired of hunting through public-library catalogs one title at a time. It’s offered free because the need is real and widely shared.
Keeping it free
Running the site costs a little, and the person behind it isn’t wealthy. If the tool saves you time, a small tip genuinely helps; modest, unobtrusive ads cover the rest. Either way, the tool stays free for the librarian who needs it.
Open the tool