How to find profitable public domain audiobooks to narrate (2026)

Published 2026-05-20

Recording a public domain book is one of the few ways to produce a commercial audiobook without paying an author or rights holder a cent up front. The text is free to use, so your only real costs are studio time, your own labor, and whatever editing and proofing you outsource. That changes the math on every title you consider.

But “free to record” is not the same as “worth recording.” The catalog is crowded with the same few hundred classics, narrated dozens of times each. The skill is not finding a public domain book. It is finding one that people will actually buy and that is not already buried under fifty competing versions. Here is how to do that work before you ever open your DAW.

Why public domain titles change the cost equation

When you license a contemporary book through ACX, you are either splitting royalties with the rights holder or being paid per finished hour by them, and in both cases the author or publisher controls the title. With a public domain work, nobody holds those rights. You can record it, sell it, and keep the full creator side of whatever the platform pays out.

That has two practical effects. First, your break-even point is lower, because there is no rights fee and no royalty split with an author baked into the deal. Second, you own the production decisions: cover, casting, pacing, whether you do a straight read or a fully voiced performance. The trade-off is that anyone else can record the same book too, which is exactly why selection matters so much.

How to verify public domain status (and why you must confirm it yourself)

This is the part to get right, because the cost of being wrong is a takedown notice or worse. Treat the guidance below as a rough orientation, not legal advice, and confirm every title yourself before you record.

In the United States, the most reliable rule of thumb in 2026 is the pre-1929 cutoff. Works published in the US before 1929 are generally in the public domain. Each January the cutoff rolls forward by one year, so the window keeps widening. Separately, the broader framework for newer works is life of the author plus 70 years, but that calculation is harder to apply by hand and has many exceptions, so the published-date cutoff is the safer first filter for most narrators.

A few things that trip people up:

• Status varies by country. A book that is public domain in the US may still be under copyright in Canada, the UK, the EU, or elsewhere, and the reverse is also true. Confirm the status in the territory where you intend to sell, not just where you live.
• Translations and editions carry their own rights. An old novel may be public domain, but a 1990s translation of it, or a specific annotated edition, may not be. Use the original text, not a modern reworking.
• Public domain in the source does not mean clearance to use someone’s recording. A LibriVox reading or a scanned edition can have its own terms. You are recording your own performance from a free text, not reusing another person’s audio.

When in doubt, check the title against more than one source, look at the actual publication date of the edition you are working from, and if real money is on the line, run it past someone qualified to advise on copyright in your selling territory. The few minutes this takes is cheap insurance.

Where clean public domain texts come from

You want a text that is accurate and easy to read from, because a sloppy source slows your session and creates pickups. Two sources cover most needs:

• Project Gutenberg is the largest archive of public domain books and the default starting point. Coverage is enormous, though formatting can be uneven and some files carry transcription artifacts from older scans.
• Standard Ebooks takes public domain texts and reformats them to a clean, consistent, modern typographic standard. The catalog is smaller, but the files are noticeably nicer to narrate from, with fewer odd characters and better structure. When a title you want is available there, it is usually the better reading copy.

For either source, read a few pages before committing. Archaic spelling, heavy dialect, or dense footnotes can turn a short book into a long, painful session.

How to gauge demand against competition on ACX

This is where most of your research time should go. A profitable title sits in the gap between “people want it” and “not many people have recorded it.”

Work through it like this:

1. Count the existing narrations. Search the title on Audible and ACX and tally how many audiobook versions already exist. One or two is promising. A dozen or more, as you will see with the most famous classics, means you are walking into a price-and-attention fight you probably lose as a newer narrator.
2. Read the sales rank as a demand signal. Audible best seller rank is a rough, relative indicator, not a precise sales figure, but it helps you compare titles. A book with a healthy rank and few audio versions is the pattern you are hunting for. Treat the number as directional only.
3. Avoid the obvious traps. Pride and Prejudice, Dracula, The Adventures of Sherlock Holmes, and a short list of other classics already have dozens of excellent, well-reviewed recordings, several by name narrators. The text is free, but the shelf is full. Demand is real, yet your version is unlikely to surface.
4. Look at the reviews on existing versions. If the current recordings are thin, poorly produced, or complain-about-the-narrator territory, that is an opening. A genuinely better performance of a wanted book can earn its place even against existing competition.

Picking underserved niches

The repeatable opportunities tend to cluster in a few places:

• Lesser-known works by famous authors. Everyone has recorded the one or two signature novels. The same author’s minor novels, essays, letters, or early work are often wide open while still carrying name recognition that helps discovery.
• Short story and essay collections. These are quicker to produce, easy to sample, and frequently underserved compared with the headline novels.
• Niche nonfiction. Old works on a specific craft, history, philosophy, regional travel, or a defined hobby can find a small but motivated audience with almost no audio competition. A narrow, well-served niche often beats a broad, crowded one.
• Genre fiction outside the canon. Early science fiction, mystery, adventure, and horror beyond the few famous titles can have devoted listeners and a thin audio catalog.

The thread running through all of these: pair real listener interest with a shelf that is not already full.

The royalty math, at a high level

Treat every number here as approximate. Platform terms change, and your results depend on price, reviews, and discovery, so model conservatively.

There are two broad ways narrators get paid on ACX. With a royalty share arrangement you take an ongoing percentage of sales rather than an up-front fee, which means your income scales with how the title actually sells over time. For a public domain book you produce on your own, you effectively hold the full creator side of that royalty rather than splitting it with an author. The other common model on the platform is per-finished-hour, a flat amount the rights holder pays you for each finished hour of audio, which front-loads your income and removes your exposure to how the title sells later.

For self-produced public domain titles, your economics come down to a simple comparison: the hours of work to produce the book, against the royalties you expect that specific title to earn given its demand and competition. A short, well-chosen book in an underserved niche can pay back its production hours faster than a long, crowded classic that sells in tiny volumes against a dozen rivals. Run the estimate per title, keep your assumptions modest, and remember that no title is guaranteed to sell.

A simple selection checklist

Before you commit a title to your production schedule, confirm:

• Copyright status verified in your selling territory, from more than one source, using the original text and edition.
• A clean reading copy located (Standard Ebooks if available, otherwise Project Gutenberg), and read-tested for dialect, footnotes, and archaic spelling.
• Existing audio versions counted, with a manageable level of competition.
• Sales rank checked as a directional demand signal.
• Reviews on existing versions read for quality gaps you can beat.
• Production hours estimated and compared against realistic, conservative royalty expectations.
• A clear reason this title is underserved (lesser-known work, niche nonfiction, weak existing recordings).

If a title clears all seven, it is a candidate worth recording. If it stalls on the first one, stop and confirm rights before anything else.

Where Build Bench fits

The slow part of all this is the research: working through dozens of candidate titles, counting competing narrations, reading sales signals, and finding the underserved corners worth your studio time. That is what audiobooks.buildbench.ca does. It curates public domain audiobook opportunities and ships them with competition counts and voice-match notes so you can skip to the shortlist.

The tiers are, in CAD:

• Spotter, about $29, a single curated title with its competition and voice-match notes.
• Catalog, about $99, a batch of 25 vetted opportunities to fill out a production pipeline.
• Production briefs, about $199, deeper write-ups to plan around.

Treat it as a research time-saver, not a shortcut around due diligence. We do the legwork on demand and competition, but you still verify the copyright status yourself in your selling territory, and you still do the recording. We make no promise that any title will sell. What we can do is hand you a cleaner starting list than scrolling Audible search results one classic at a time.