Can Library Catalogs Show How Many Books Are In The World?

Counting all the books in the world is a deliciously nerdy idea, but the short, enthusiastic truth is: library catalogs can help us estimate, they can show what libraries collectively hold, but they can't definitively tell you how many books exist on Earth.

Library catalogs are essentially inventories for collections — a single library’s catalog lists what that library owns or subscribes to, and union catalogs like OCLC’s WorldCat aggregate many libraries’ records into one gigantic pool. Those union catalogs are amazing research tools and they contain hundreds of millions of bibliographic records, representing many editions, translations, formats, and media. But a huge catalog still reflects the limits and quirks of what was cataloged, what libraries acquired, and how different editions were recorded. That means duplicate records, varying metadata quality, and spotty coverage for self-published works, small-press runs, pre-ISBN items, and non-Western publishing can all skew any attempt to use catalogs as a global counter.

Where it gets really sticky is in definitions. Are you counting unique works (like Tolstoy’s 'War and Peace' as one), unique editions and translations, physical copies, or every distinct manifest form (paperback, hardcover, e-book, audiobook)? Different projects choose different definitions. ISBN-based tallies are convenient but miss older books, many academic theses, small-press zines, and anything published without an ISBN. Legal-deposit national bibliographies are thorough for what falls under their laws, but they only cover a country’s legal-deposit scope and won’t capture everything printed or self-published there. Then there’s duplication: one library’s record for an edition might be separate from another library’s record for that same edition, and deduplication across millions of records is an imperfect science.

People trying to estimate the total have combined sources — union catalogs, ISBN agency records, national bibliographies, and web-scale crawls like Google Books — and come up with different figures. For instance, Google’s 2010 estimate of distinct books got a lot of attention (it was roughly in the low hundreds of millions), while WorldCat and similar services often point to hundreds of millions of records overall. None of these numbers is a single, definitive “book count,” though; they’re snapshots informed by scope, definition, and technical limits. For me, that uncertainty is part of the charm: there’s something thrilling about the idea that despite centuries of cataloging we still can’t pin down the full extent of human storytelling and knowledge. It’s messy, fascinating, and makes me want to dive into catalogs and special collections just to see what surprising titles turn up.