Why ACX Rejected Your Audiobook (and How to Fix Each Problem)

If you have just received an email stating that your project has failed quality assurance, that is a rotten email to open. You thought you were finished, and now you are not. The thing to hold onto is that an ACX rejected audiobook is almost never a verdict on the book or on the performance. Most rejections come down to a handful of technical problems, and each one has a specific fix.

ACX's quality assurance team reviews submitted files within 10 business days, according to the official ACX producer's checklist. When a title does not pass, it is usually because the files deviate from their specific technical standards. Below are the reasons ACX sends audiobooks back, with the fix for each, followed by a routine to run before you resubmit so you only go through the queue once more.

Start With the Numbers ACX Checks

Most rejections are technical: a file misses one of the published specs. Keep the official ACX audio submission requirements page open in a tab while you work. Most of what is on it is a pass or fail measurement, not a judgment call.

For example, the requirement for your noise floor is that it must be less than -60dB RMS. If your recording environment has a noticeable hum from a computer fan or an air conditioner, you will likely fail this check. Similarly, your RMS levels must sit between -23dB and -18dB, and your peaks must stay below -3dB. If you are struggling to keep your audio consistent, it helps to review our guide on audiobook file formats and specs for Audible to ensure your export settings are locked in correctly.

Common Reasons for Rejection and How to Fix Them

Improper File Grouping and Chapter Formatting

According to the ACX blog, the most frequent reason for rejection is improper file grouping. You must provide exactly one chapter or section per file. If you have lumped multiple chapters together, or if your files are cut mid-sentence, the quality assurance team will flag it immediately. ACX expects a clean, logical structure where every file represents a distinct, standalone unit of your manuscript.

To fix this, you must go back into your project and split your tracks so that every file represents a single, self-contained section. If you are using Audie to generate your audio, you will find that the system automatically handles the chaptering for you. It ensures each exported MP3 file is its own distinct chapter, which saves you a significant amount of manual labor and prevents those common grouping errors.

Noise Floor and Volume Issues

If your audio is too quiet or too loud, it falls outside the required -23dB to -18dB RMS range. Often, this happens when an author tries to normalize their audio but misses the mark, or when they leave the audio unprocessed. On the flip side, a high noise floor (that persistent hum or hiss sitting underneath your voice) is usually caused by the recording environment: a computer fan, an air conditioner, or a noisy microphone setup.

The fix here is to use a dedicated loudness normalization tool. You want to adjust the gain so that the average volume hits that target window, and then apply a limiter to ensure your peaks never cross the -3dB threshold. If the noise floor is the culprit, do not rely on aggressive noise reduction software, as it often creates a "watery" or robotic sound. It is always better to re-record in a quieter space if you can, and to keep any noise reduction gentle.

Wrong Bitrate, Sample Rate, or a Mix of Mono and Stereo

This group of rejections has nothing to do with how your book sounds and everything to do with how the files were exported. ACX asks for MP3 files at 192 kbps or higher, encoded at a constant bit rate (CBR), with a 44.1 kHz sample rate. Many audio editors default to a variable bit rate because it makes smaller files, and that one default setting sends a surprising number of otherwise clean audiobooks straight back to their authors. Your files also need to be all mono or all stereo. A book where most chapters are mono and two are stereo will be flagged, even if every chapter sounds fine on its own.

The fix is a re-export, not a re-record. Open your export settings, choose MP3, set the bit rate mode to constant, pick 192 kbps or higher, and confirm the sample rate is 44.1 kHz. Then decide on mono or stereo and export every single file, credits included, with the same settings. While you are in there, check the length of each file too: ACX does not accept a file longer than 120 minutes, so a very long chapter needs to be split at a natural break. And make sure you have a retail sample ready, because ACX requires one and it has to be 5 minutes or less.

Missing or Incorrect Room Tone

ACX recommends between 1 and 5 seconds of room tone at the beginning and end of each file. This silence should not be absolute digital silence; it should be the natural sound of your recording space. If you cut the audio so tightly that there is zero air, it can sound jarring to the listener during chapter transitions.

You can fix this by manually adding a room tone clip to the start and end of your edited files. If you find your recording is totally silent, you can record a few seconds of ambient noise from your room, just the sound of the air in the space, and splice it onto the ends of your chapters. Just be careful that the volume of this ambient noise matches the overall volume of your narration so it does not sound like you have switched rooms halfway through the book.

Inconsistent Credits and Narration

Credits trip people up because they are easy to forget and easy to word loosely. Your opening credits must state the title, the author, and the narrator. For the closing credits, the approach ACX recommends is to say "You have been listening to", restate the title, author, and narrator, and finish with "The End." A file set with no opening or closing credits is incomplete, so record both.

Listen to both credit files against that wording. If the narrator's name is missing, or the title is read differently from how it appears on the cover, re-record just that credit and swap it in. Keep the credits as their own files, separate from the first and last chapters, so they do not create a file-grouping problem.

Policy Issues Regarding AI Narration

One rejection reason cannot be fixed in an audio editor. If you used AI to generate your narration and uploaded it through the standard ACX submission process, you may face a rejection based on policy. The official ACX requirements state that unauthorized use of AI or automated recordings is prohibited. If this is the reason for your rejection, you cannot simply edit the files to fix it; the issue is the origin of the audio itself.

I recommend reading our detailed breakdown on whether ACX allows AI-narrated audiobooks in 2026 to understand your options. Many authors find that distributing their AI-narrated books through platforms that explicitly support and disclose AI content is a much more reliable path. If you are looking for high-quality, expressive voices for your projects, you might consider using ElevenLabs, which offers some of the most natural-sounding neural voices available today. Please note that this article contains affiliate links, and I may earn a commission if you use them.

Which Fixes Need a New Recording and Which Only Need a New Export

Before you book more studio time or regenerate a whole book, sort the items in your rejection email into two piles. Most of them land in the cheaper one.

Problem	What it takes to fix
Variable bit rate, bit rate under 192 kbps, wrong sample rate	Re-export with the correct settings. No new audio needed.
Mix of mono and stereo files	Re-export every file in one channel format.
More than one chapter in a file, or a file over 120 minutes	Split at chapter or section breaks and re-export.
Volume outside -23dB to -18dB RMS, or peaks above -3dB	Normalize and limit the existing recording, then re-export.
Missing room tone at the start or end	Add a short clip of room tone to the file ends.
Missing or incomplete credits	Record the credit only, as its own short file.
Noise floor above -60dB RMS	Try gentle noise reduction first. If the audio turns watery, re-record in a quieter space.
Skipped or repeated lines	Re-record and patch the affected passages.

Work through the export-only pile first. It is quick, and you will often find that a single wrong export setting was applied to every file, which means one corrected export clears most of the list at once.

The Pre-Resubmission Routine

Before you hit the submit button again, you need to be absolutely certain that your files are ready. The first step is to run every file through ACX Audio Lab. This is a free checker provided by ACX that automatically measures your files for RMS, peak levels, bitrate, bitrate method, sample rate, mixed channels, duplicate files, and spacing. It catches the measurable problems that are easy to miss by ear. If a file fails Audio Lab, fix it before you submit: the human review will send it back for the same reason.

However, keep in mind that the Audio Lab does not check for everything. ACX says plainly that it does not detect noise floor problems or editing errors, and it will not tell you whether your credits are worded correctly. After running the tool, I strongly suggest you do a manual listen-through of at least the opening and closing of each file. Verify your credit wording, ensure your room tone is present, and make sure that your chapters are correctly ordered. If everything looks and sounds correct, you are ready to upload your audiobook to Audible step by step once more.

Frequently Asked Questions

Does passing the ACX Audio Lab guarantee approval?

No, it does not. While the ACX Audio Lab is excellent for checking technical specs like RMS and peak levels, the human quality assurance team will still review your files for other issues, such as noise floor, credit wording, and narration consistency. Use the tool as a baseline, but always perform your own manual quality check before submitting.

How long does it take for ACX to review my resubmission?

ACX says its quality assurance team reviews files within 10 business days of submission, so it is sensible to plan on a similar wait after you resubmit corrected files. ACX does not publish a faster lane for resubmissions, which is the best argument for fixing everything in one pass.

Can I use noise reduction software to fix a high noise floor?

You can use gentle noise reduction, but be very careful. If you over-process your audio, it will start to sound watery or robotic, which is another reason for rejection. It is always better to record in a quiet space than to try and save a noisy recording with heavy software filters that degrade the natural quality of your voice.

What should I do if my chapters are split incorrectly?

The only way to fix this is to manually re-export your audio. You must ensure that each file contains exactly one chapter or section. If you have chapters combined, you need to go back to your editing project, create separate tracks for each chapter, and export them as individual files that meet the standard naming and formatting requirements.

Should I check my narration against my manuscript?

Yes. Skipped paragraphs and repeated lines are editing errors, and ACX Audio Lab will not find them for you. Even when every technical spec passes, a proof-listen against the manuscript is the only way to catch them, so it is time well spent.

If you would rather skip the file-splitting and export headaches on your next project, you can create a chaptered audiobook with Audie, where every chapter comes out as its own MP3 file. Have a look at Audie pricing to see what fits your book, and check the full ACX audio submission requirements checklist before you upload anywhere. Just remember what I said above about ACX and AI narration: for an AI-narrated audiobook, a distributor that accepts disclosed AI narration is usually the realistic home.