Why Your EPUB Won't Open (and How to Fix Every Common Error)

If your reader app crashes the instant you try to import the EPUB, or the file shows as 0 bytes in the Files app, the ZIP container itself is physically broken. This is the only category where the file is genuinely unrecoverable without a clean copy—every other failure below can be repaired.

**Quick diagnosis:** Rename the `.epub` to `.zip` and try to open it. On iPhone, tap the renamed file in Files. On desktop, use 7-Zip or The Unarchiver. If the ZIP utility cannot list the contents, the archive is damaged beyond what any EPUB tool can fix.

**Fix:** Re-download the file from the original source. If no clean copy is available, desktop ZIP repair tools (like DiskInternals ZIP Repair or WinRAR's repair function) can sometimes recover the contents. Once the ZIP structure is intact, open the EPUB in EpubTidy to validate and repair the internal package.

The EPUB specification requires two structural files at specific locations: a `mimetype` file as the first entry in the ZIP (uncompressed, containing exactly `application/epub+zip`), and a `META-INF/container.xml` that points to the OPF package document. If either file is missing, damaged, or stored in the wrong order, strict readers reject the file.

This commonly happens when someone creates an EPUB by manually zipping files without following the mimetype-first requirement, or when a ZIP utility recompresses the archive and moves the mimetype entry. Some readers are lenient (Apple Books often ignores this), but Kindle Send-to-Kindle and Kobo are strict.

**Fix:** EpubTidy re-packs the archive with the correct mimetype entry when you export the book. On desktop, EPUBCheck identifies the specific error, and you can use Sigil to repackage the EPUB correctly.

The OPF file (`content.opf` or `package.opf`) is the master registry of every file inside the EPUB: HTML chapters, images, CSS, fonts, and the reading order (spine). If the manifest references a file that does not exist, omits a file that does exist, or contains invalid XML, the reader may display a blank book, skip chapters, or refuse to open entirely.

Common causes include manual editing that introduced a typo in a filename, conversion tools that failed to register all images, or merging chapters from multiple sources without updating the manifest.

**Fix:** EpubTidy's inspection mode identifies orphaned files (present in the ZIP but not in the manifest) and phantom references (listed in the manifest but missing from the ZIP). The repair function reconciles both directions automatically.

Mojibake—garbled text where accented characters, em dashes, curly quotes, and non-Latin scripts display as garbage symbols—is caused by a character encoding mismatch. The EPUB's HTML files are encoded in one format (commonly UTF-8 or Windows-1252) but declare a different encoding, or declare no encoding at all.

This frequently affects EPUBs converted from older Word documents, PDFs, or legacy formats. The converter may output Windows-1252 bytes but write `<meta charset="UTF-8"/>` in the HTML header, causing readers to interpret every non-ASCII byte incorrectly.

**Fix:** EpubTidy detects encoding mismatches and normalizes all HTML content to valid UTF-8 with correct declarations. On desktop, Sigil can re-encode individual files, and Calibre's conversion pipeline includes encoding detection options.

If an EPUB was purchased from a store that applies Digital Rights Management (Adobe DRM, Apple FairPlay, or Kobo's DRM), it can only be opened in the store's authorized reader app. Attempting to open a DRM-protected EPUB in any other app—including EpubTidy, Calibre, or Apple Books—will fail with a generic error.

This is not a corruption problem and no repair tool can fix it. The file is intentionally encrypted to restrict access. If you believe you should have access, contact the store where you purchased the book to reauthorize your device or download a fresh copy.

**How to tell:** DRM-protected EPUBs typically contain an `encryption.xml` file inside the `META-INF` directory. If you rename the `.epub` to `.zip` and see this file, the book is likely DRM-locked.

Sometimes a file has the `.epub` extension but is actually a different format—a renamed PDF, a MOBI file, a ZIP of HTML files without EPUB structure, or even a plain text file. Readers see the extension, attempt to parse it as an EPUB, and fail.

This happens when download services mislabel files, when users manually rename extensions, or when a server serves the wrong Content-Type header and the browser saves the file with an incorrect extension.

**Diagnosis:** Rename the file to `.zip` and inspect the contents. A valid EPUB contains a `mimetype` file, a `META-INF/` directory, and at least one OPF package document. If these are missing, the file is not an EPUB regardless of its extension.

Very large EPUBs—particularly those containing hundreds of high-resolution images, embedded fonts for multiple languages, or extremely long single-chapter HTML files—can exceed the memory limits of mobile reading apps. The app may crash, display a spinning indicator indefinitely, or load only the first few chapters.

Kindle's Send-to-Kindle service rejects files larger than 50 MB. Apple Books handles larger files but may struggle with EPUBs containing individual HTML chapters over 300 KB (the IDPF recommendation).

**Fix:** If the EPUB contains unnecessarily large images, compress them before embedding (JPEG quality 80 is usually sufficient for reading). If a single HTML chapter is enormous, split it into smaller files. Calibre's 'Split' plugin can automate chapter splitting for desktop users.