KDP Formatting

KDP error: EPUB doesn't pass EPUBCheck validation

TL;DR

KDP runs EPUBCheck on every Kindle upload. Validation errors — malformed XHTML, broken OPF references, missing required metadata, unsupported media — trigger rejection. Run EPUBCheck locally, fix the reported errors line by line, and re-upload. Our free /audit/kdp-readiness/ Score includes EPUBCheck output for free.

Last reviewed by Robert Prime — May 2026

Quick Answer: Your EPUB file fails KDP's automated validator. Run it through EPUBCheck (free) at epubcheck.org or via Pagina's free desktop tool to identify the specific errors. Most common: malformed metadata, broken NCX/nav, missing manifest entries, or unclosed HTML tags.

Full reasoning, why-it-happens, and the exact fix below.

UK note: UK-specific considerations apply — ISBN purchases go through Nielsen (not Bowker), VAT rules differ from the US (print books are zero-rated; ebooks carry 20% VAT), and GDPR applies to any email/customer data. See our UK self-publishing guides for specifics.

We see this come through our formatting queue at publishing.co.uk regularly, so the patterns and fixes here are based on what actually works at upload.

What this error means

EPUBCheck is the open-source validator maintained by the W3C and IDPF that confirms an EPUB file conforms to the spec. KDP runs it on every Kindle upload. If EPUBCheck reports errors (not just warnings), the file is rejected with "Your EPUB file did not pass validation."

The error report lists each issue with file, line number, and severity (FATAL, ERROR, WARNING). Only FATAL and ERROR cause rejection; WARNINGs are allowed.

Common errors:

  • Missing required metadata (title, creator, language, identifier)
  • Malformed XHTML (unclosed tags, invalid attributes)
  • Broken references in the OPF manifest
  • Files referenced but not included in the package
  • Files included but not declared in the manifest
  • Unsupported MIME types (e.g. TIFF images, which Kindle rejects)
  • Encoding issues (non-UTF-8 XHTML)

📎 Source: KDP's authoritative documentation on this rule is at EPUBCheck (W3C standard validator).

Why it happens

Hand-edited EPUBs from Sigil or Calibre with malformed HTML are the top cause. Authors edit raw XHTML to tweak something and break the syntax.

Exporting from Word as EPUB via plugins (third-party converters) often produces invalid markup — orphaned <span> tags, missing alt text, broken table structure.

Calibre's conversion engine can produce technically valid but slightly off EPUBs depending on input. Word .docx → EPUB conversion is the most error-prone path.

Renaming files inside the EPUB without updating the OPF manifest. The manifest is an XML inventory of every file; rename one and the reference breaks.

Mixing EPUB 2 and EPUB 3 metadata — Atticus and some online converters output a hybrid that EPUBCheck flags.

Missing language declaration in the OPF (<dc:language>en-GB</dc:language>). Surprisingly common because some tools omit it.

Image format mismatch — referencing a .tif file in HTML but KDP requires JPEG/PNG/GIF/BMP only.

The fix

Step 1: Run EPUBCheck locally. Download the latest from github.com/w3c/epubcheck — it's free and open source. You need Java installed.

Step 2: From a terminal, run:

java -jar epubcheck.jar your-book.epub

Step 3: Read the output. Each error has the form:

ERROR(RSC-007): your-book.epub/OEBPS/content.opf(4,42): Referenced resource missing in the package.

Translation: line 4, column 42 of content.opf refers to a file that isn't in the EPUB.

Step 4: Open the EPUB with Sigil (sigil-ebook.com, free). Sigil treats the EPUB as a project with all files visible — XHTML, CSS, OPF, NCX, images.

Step 5: Fix each error in order:

  • "Referenced resource missing" → add the missing file to the manifest or remove the reference
  • "Missing required metadata" → open content.opf and add the missing dc:title / dc:creator / dc:language / dc:identifier entries
  • "Element name does not match" → close the unclosed tag in the XHTML file at the reported line
  • "Image format not allowed" → convert TIFF/WebP images to JPG, replace, update references
  • "Encoding not UTF-8" → in Sigil, right-click the file → Encoding → UTF-8 without BOM

Step 6 (Atticus / Vellum users): Both export valid EPUB 3 in most cases. If you're failing, regenerate the EPUB rather than hand-editing — the export is usually fine but a previous manual edit broke it.

Step 7 (Calibre users): Calibre → Edit Book → Tools → Check Book runs an internal validator. Fix the issues it lists. Re-save.

Step 8: Re-run EPUBCheck. Repeat until you see "No errors or warnings detected" or only warnings.

Step 9: Upload to KDP. Their preview tool will render the corrected EPUB on Kindle Fire.

How to pre-flight it

Our free KDP Readiness Score runs EPUBCheck on your file, parses the output, and shows each error with file plain-English fix suggestion. We also check the 30+ other KDP rules including image format, cover dimensions, and metadata.

FAQ

Are warnings as bad as errors? No — EPUBCheck warnings don't trigger KDP rejection. Only FATAL and ERROR do. Warnings are best-practice suggestions.

Can I upload my .docx instead? Yes — KDP accepts .docx and converts it internally. But you have less control over the result. EPUB is preferred for serious authors.

What about KPF or KFX format? KDP accepts EPUB and converts to KFX server-side. You don't need to generate KFX yourself unless you use Kindle Create.

My EPUB passes locally but fails on KDP — why? KDP runs a stricter EPUBCheck plus their own additional checks (cover dimensions, image format, file size). Cross-check against their full guidelines, not just EPUBCheck.

Frequently asked questions

Is EPUBCheck the same as KDP's validator?

Very similar — both implement the EPUB 3.0 specification. If your file passes EPUBCheck, it'll almost certainly pass KDP's validator.

How do I run EPUBCheck on my file before uploading?

EPUBCheck is the same validator KDP runs internally. The official version is at w3.org/publishing/epubcheck/, and most authoring tools (Sigil, Calibre) integrate it. Our free KDP Readiness Score runs it on your uploaded EPUB and shows you every error in plain English.

Will publishing.co.uk fix EPUB validation errors for me?

Yes — EPUB validation errors usually come down to malformed HTML inside the EPUB or missing manifest entries. We fix the underlying file (most errors are 5-10 minute corrections), re-validate, and re-deliver. From £69 turnaround typically 24 hours.

Are validation errors just warnings or do they actually fail KDP review?

Errors fail KDP review and are not negotiable. Warnings sometimes pass but are worth fixing — they often indicate the file will render oddly on certain Kindle devices even when the upload succeeds.

If you got the rejection above, you may also want to check these related issues — they tend to cluster:

Full list: KDP formatting errors hub · KDP rejection fixes hub · KDP rejected my book


About this guide

This page is part of a series of UK-focused KDP rejection guides at publishing.co.uk, each documenting a specific reason KDP can reject a print or Kindle file and the exact fix. Written by Robert Prime — founder of publishing.co.uk, co-runs the LoveReading network, and has overseen 500+ KDP submissions through formatting work in this category (epub).

If you'd rather have someone else handle this and the other 35 issues KDP checks for, our formatting service is from £69 with a 3-day turnaround and a 100% KDP-acceptance guarantee.

Run a free KDP Readiness Score on your file before you upload — catches this issue and 35 others, in 60 seconds.

Free · 60 seconds · No payment

Don't risk a KDP rejection — score your file first.

Drop your DOCX, PDF or EPUB and we run the same 30+ checks Amazon does — margins, gutter, image DPI, font embedding, ToC, blank pages, ISBN match, bleed — and score it /100 with the exact rejection risks flagged.

Score my file →
Robert Prime

Robert Prime

Robert Prime is a best-selling self-published author, veteran eCommerce strategist, and the founder of publishing.co.uk.

Robert Prime — Founder of publishing.co.uk

About the Author

Robert Prime

Robert Prime is the founder of publishing.co.uk and a co-owner of LoveReading.co.uk. A Forbes Business Council member with 25+ years in eCommerce, he writes about Amazon KDP strategy, scaling indie author businesses, and the commercial side of self-publishing.

Reading about KDP? Score your file free in 60 seconds. Score my file →