Find the failing path
The scan names the package path, manifest entry, anchor, or identifier behind the upload blocker.
KDP upload failure · Kindle Previewer message · EPUBCheck code
Upload the exact .epub that failed. The report names the affected path, separates checked repairs from manual work, and only offers a repaired copy after the file passes the checks again.
What the report answers
Preview a sample reportDownload safety gate
If either check still reports a problem, the report explains the next step but no download appears.
The scan names the package path, manifest entry, anchor, or identifier behind the upload blocker.
A repair is written only when the correct package change is deterministic and does not require a book decision.
The separate copy becomes downloadable only after both post-repair checks pass.
Repair limits
This is not a general-purpose EPUB service. It only writes changes when the correct structure fix is clear and verifiable.
Common KDP blockers
Compare your code or message with these common blockers. If none matches, open the full error guide index.
Use this when EPUBCheck reports duplicate ids after export or editing, especially around footnotes, headings, backlinks, or TOC anchors.
Use this when KDP or EPUBCheck says an image, stylesheet, chapter, table of contents file, or other referenced file is missing.
Use this when a link points to a note, heading, anchor, or table of contents target that no longer exists.
Use this when the EPUB package points to a book identifier that is missing from the metadata.
Use this when the EPUB was zipped incorrectly, often after manual zipping or archive-tool changes.
FAQ
No. It only fixes cases where the correct change is clear. If the fix would require choosing chapter order, rewriting text, picking a heading, or redesigning layout, the report stops at diagnosis.
Only when the file is supported, the change is automatic, and the repaired EPUB passes the structure scan and EPUBCheck again. Retry the downloaded file in KDP or Kindle Previewer before treating it as final.
No. Upload an existing EPUB 2 or EPUB 3 book. EPUB Fixer does not turn DOCX, PDF, KPF, KCB, ZIP archives, fixed-layout books, comics, or image-first books into EPUB.
Yes. You can scan an EPUB for free. If a checked repaired copy is available, the result page shows a download while the upload is still available.
EPUB Fixer works best with normal text-first EPUB 2 and EPUB 3 books. It does not handle fixed-layout books, comics, DRM-protected files, KPF, KCB, complex multimedia, or heavily scripted EPUBs.
EPUBCheck tells you which part failed its check. EPUB Fixer turns that message into a next step and only offers a repaired copy when one deterministic structure change passes the structure scan and EPUBCheck again.
Yes when the issue is a clear EPUB structure problem. Upload the EPUB, check the report, and download a repaired file only if the automatic repair passes both post-repair checks. Missing content, layout, cover design, and conversion still need another workflow.
KDP and Kindle Previewer are the main focus. If another upload checker shows the same EPUBCheck-style structure error, use the report to compare the affected path before changing the file.
Reading apps such as Calibre can be forgiving. KDP, Kindle Previewer, and EPUBCheck are stricter about package files, navigation, anchors, media paths, metadata, and ZIP packaging.
No. The scan does not modify your original EPUB. When one deterministic repair is available, EPUB Fixer creates a separate copy without rewriting prose, redesigning pages, or guessing what the book should say.
No. The original EPUB, repaired copy, and temporary scan files are deleted automatically after 24 hours. You can delete those files sooner from the result page; scan metadata and saved follow-up records have a separate lifecycle explained on the Privacy page.
Upload one .epub file that is 30 MB or smaller. Larger files, ZIP archives, PDFs, DOCX files, and source project files are rejected before scanning.
Ready to retry?