Where to download public domain books

Start here: Standard Ebooks if the book you want exists there, Project Gutenberg for everything else, and the Internet Archive when only a scan survives.

The main sources

• Project Gutenberg. The original and still the largest, with tens of thousands of titles in EPUB and plain text, plus an OPDS catalogue you can browse from inside a reading app. Editions are plain, and quality varies with how old the transcription is.
• Standard Ebooks. A volunteer project that takes public-domain texts, proofreads them, and typesets them properly: real small caps, sensible chapter structure, modern semantics. The catalogue is much smaller because each book takes real work. If your book is in it, take that edition.
• The Internet Archive and Open Library. Enormous, including scans of books that were never retyped. The reading experience is often a PDF of a scan, which is why EPUB vs PDF matters here.
• Wikisource. Community transcriptions with source-text discipline, strong on documents, speeches and out-of-copyright non-fiction.
• HathiTrust and university libraries, for scholarly and older academic material.
• National and language-specific collections. Gallica for French, DBNL for Dutch literature, Faded Page for Canadian public domain, and similar projects in most languages. These are where you find work the English-language sites never touched.

For audio, LibriVox is the equivalent: volunteers reading public-domain books, free to download and keep. More on building from it in how to build your own audiobook library.

Public domain is a local question

Copyright expires on different schedules in different countries, so a book can be free in one place and protected in another. Broadly, much of Europe and the United States use the author’s life plus seventy years, while a number of countries use life plus fifty, and the United States has separate rules for older published works that move forward one year at a time.

What that means practically: a site hosting works legally in Canada may host something still in copyright in Germany. Reputable projects say which rules they follow. If you are unsure about a specific title, the safest check is your own national library’s guidance.

Getting them into your reader without the browser dance

Downloading in a browser, finding the file, then importing it is the tedious part. Two ways around it:

1. OPDS catalogues. Gutenberg and Standard Ebooks both publish one, so a reading app can browse and download directly. Explained in what is OPDS. Lectern has both built in.
2. Download to your books folder rather than to Downloads, so the file is already where your library lives. See organising a digital library.

Picking a good edition

Two files of the same novel are not the same book. Before committing to a long read, check three things: does the table of contents work, are chapter headings real headings rather than centred bold text, and does a page of dialogue look right. A badly converted 1990s transcription will annoy you for six hundred pages.

Where several editions exist, prefer the one with a named source and a recent revision date. Where only a scan exists, expect to read it as a PDF or to spend time on OCR.

What to do with them

Keep them. These files have no DRM, no account, and no expiry, which makes them the most durable part of any digital library. They’re also the easiest way to test a reading setup: pick a novel you know well, load it, and see whether the app treats the typography kindly before you buy anything.

Frequently asked