How to fix a corrupted Word document

What a corrupted Word document looks like

Corruption shows up in a handful of ways. Word reports 'We found a problem with its contents', 'Word found unreadable content' or 'The file is corrupt and cannot be opened'. Sometimes the document opens but shows boxes or a wall of garbage characters, Word hangs at 'Opening', or the file is listed at 0 KB.

The cause is usually an interrupted save: the laptop ran out of battery, a USB stick was pulled, the network dropped, or Word crashed mid-write. Less often a corrupted picture inside the file is the problem, or the file was never a Word document at all. The steps below go from the gentlest to the most drastic.

Before you touch anything, make a copy of the file and work on the copy. Every repair that rewrites the file can destroy data the next method would have recovered.

Quick checks before you repair anything

If any of these opens the document, save it at once under a new name and you are done. If none of them helps, the Repair Word tool further down rebuilds the file in the browser.

• If the file is on a USB stick, network share or cloud folder, copy it to the Desktop first and open the copy. Many 'corrupted' files are just unreachable.
• Check the size in File Explorer or Finder. A file of 0 KB has no content left; skip straight to the backup section.
• Check the extension. A file renamed from .doc to .docx, or a PDF saved with a .docx name, is not corrupted; see the misnamed files section below.
• Try another program: LibreOffice Writer, Google Docs (upload to Drive) and Word Online are each more tolerant of damaged files than desktop Word.

How to use Word's Open and Repair

Word has a hidden repair mode that rebuilds the file structure and drops the parts it cannot parse. It keeps formatting, pictures and tables when it can.

On a Mac, the Open dialog has no repair option. Use LibreOffice, Word Online or the online repair below, or copy the file to a Windows PC.

1. Open Word. Go to File, Open, then Browse (do not double-click the file in Explorer).
2. Navigate to the file and click it once to select it, not to open it.
3. Click the small arrow on the right side of the Open button and choose Open and Repair.
4. Word opens the document, possibly with a message that it made repairs or that some content was lost. Read it, then go to File, Save As and save a new copy.

Recover Text from Any File

When Open and Repair fails, the next step gives up on formatting and pulls out whatever text it can find. You lose pictures and tables, and some junk may appear at the end, but the words are usually there.

If the text also exists in a PDF you sent earlier, PDF to Word gives a better starting point, with layout and tables kept.

1. Go to File, Open, Browse.
2. In the file type list next to the file name box (it says All Word Documents), choose Recover Text from Any File (*.*).
3. Select the file and click Open.
4. Scroll through the result. Delete the garbage at the start and end, then save as a new .docx.

A .docx is a ZIP file: look inside it

A .docx is a ZIP archive containing XML files and a media folder. You can inspect it with tools you already have, which tells you how badly damaged it is and often lets you rescue the text.

If document.xml is intact but Word still refuses the file, the error is usually elsewhere: a bad entry in the content types list, a corrupt header or a theme. Create a fresh blank .docx, rename it to .zip, replace its word/document.xml with yours, and rename back to .docx. That transplant fixes a surprising number of files.

1. Make a copy of the file and rename the copy from name.docx to name.zip. Confirm the change when Windows or macOS warns you.
2. Double-click the ZIP to extract it. If the archive itself will not open, the container is damaged; use the online repair or Recover Text instead.
3. Open the word folder. document.xml holds the whole text of the document. The media folder holds every picture, which you can copy out as ordinary image files.
4. Open document.xml in a text editor (Notepad, TextEdit or VS Code). Your text is inside the <w:t> tags. Copy the paragraphs you need into a new document.

How to repair a Word document online

If you would rather not dig through XML, Repair Word does the rebuild for you. It takes a .docx that Word refuses to open, rebuilds it and saves a clean copy you can download. It is free, no sign-up, no watermark, runs in the browser on a phone or computer, and the file is processed in memory and deleted after 30 minutes.

If the document opens but pictures are missing or broken, Extract images from Word saves every picture inside the original at its stored quality, so you can put them back.

1. Open the Repair Word page and upload the damaged .docx (work on a copy, as always).
2. Wait for the rebuild and download the result.
3. Open the repaired file in Word and check the end of the document, the tables and the pictures, which are most often affected. Save it under a new name.

Check AutoRecover, OneDrive and backup copies

A repair is only needed if no good copy exists. Often one does.

Restoring yesterday's version and redoing an hour of edits is nearly always faster than fighting a corrupted file.

• AutoRecover: in Word, go to File, Info, Manage Document, Recover Unsaved Documents. After a crash Word also shows a Document Recovery pane; save every version it lists before closing it.
• AutoRecover folder: on Windows, paste %AppData%\Microsoft\Word\ into File Explorer and look for .asd files. On a Mac, look in ~/Library/Containers/com.microsoft.Word/Data/Library/Preferences/AutoRecovery. Rename an .asd file to .docx and open it.
• OneDrive and SharePoint: right-click the file in the browser or in the OneDrive folder and choose Version history. Each save is kept as a version you can open or restore.
• Windows File History and Previous Versions: right-click the file, Properties, Previous Versions. On a Mac, Time Machine.
• Email: search Sent and Drafts for the file name. Last week's attachment is a complete copy of last week's version.

When the file is not really a .docx

A good share of 'corrupted' documents are fine files with the wrong extension. Word trusts the extension, so a PDF named report.docx fails with a contents error. You can tell what a file really is by opening it in a text editor and reading the first characters.

A .pages file from Apple Pages is a ZIP too but not a Word document; export it from Pages instead. The Convert document format tool handles DOCX, DOC, ODT, RTF, TXT, HTML and PDF.

• PK at the start: a real ZIP container (.docx, .xlsx, .pptx or .odt). If Word fails, the problem is inside the archive; use the methods above.
• %PDF at the start: it is a PDF. Rename it to .pdf, or run it through PDF to Word if you need to edit it.
• {\rtf at the start: a Rich Text file. Rename to .rtf; Word opens it directly.
• Two strange characters followed by binary data, often shown as ÐÏ: an old Word 97-2003 .doc. Rename to .doc, or upgrade it with DOC to DOCX.
• <html or <!DOCTYPE at the start: a web page. Rename to .html; Word opens it and can save as .docx.

How to prevent corrupted Word documents

Most corruption is avoidable. These habits cost nothing.

• Work on files stored on the computer or in a synced OneDrive or Google Drive folder, not directly on a USB stick or a network share. Copy to the stick when you are done, and eject it properly.
• Turn on AutoSave (the switch in the top left of Word) for files in OneDrive, and set AutoRecover to every 5 minutes under File, Options, Save.
• Save a dated copy at milestones, such as report-v3-11-09.docx.
• Keep documents a sensible size. A huge .docx full of uncompressed photos is more likely to fail during a save; DocGopher's Compress Word file shrinks the pictures inside without changing the text.
• Export a PDF of every final version (File, Save As, PDF). If the .docx is lost, the PDF converts back to an editable document.