How to Fix File Conversion Encoding in Microsoft Word

Reopen the file and pick the encoding in the File Conversion box: Word shows a preview, so you can see which choice makes the text readable before you commit. To make Word ask every time, turn on Confirm file format conversion on open under File > Options > Advanced.

These steps follow Microsoft’s page for Word for Microsoft 365, Word 2024, 2021, 2019 and 2016 on Windows.

Convert File or File Conversion: which box are you looking at?

Word can show two different dialog boxes when a text file is involved, and their names are close enough to be confusing.

  • Convert File asks what kind of file Word is opening. It appears when Confirm file format conversion on open is turned on and you open a file that isn’t in a Word format. For a text file, the choice you want is Encoded Text.
  • File Conversion asks which text encoding to use. It has the choices Windows (Default), MS-DOS and Other encoding, a list of encoding standards, and a Preview area that shows the text as it will look.

An encoding is the rule a program uses to turn the bytes in a file into letters. When Word reads a file with a different rule from the one that wrote it, accented letters, curly quotes and non-Latin scripts come out as the wrong characters. The fix is to tell Word the right rule, which is what the File Conversion box does.

Word’s own formats (.doc, .dot, .docx, .docm, .dotx and .dotm) don’t go through these boxes. Microsoft notes that Word is based on Unicode, so those documents normally don’t need an encoding choice at all.

Reopen the file and choose the encoding

If the file is already open and the text looks wrong, close it first and don’t save it, so the broken characters aren’t written back to the file. Then:

  1. In Word, select File > Options > Advanced.
  2. Scroll to the General section, select Confirm file format conversion on open, and select OK.
  3. Open the file again from File > Open. If you need help finding it, opening a file in Word covers the Browse dialog and recent files.
  4. In the Convert File box, select Encoded Text and continue.
  5. In the File Conversion box, select Other encoding, then select an encoding in the list.
  6. Check the Preview area. When the sample reads correctly, select OK.
Word File Conversion dialog with Other encoding selected and Unicode (UTF-8) chosen, with a correct preview
(1) Choose Other encoding, then (2) select Unicode (UTF-8) and check the preview. (Illustration)

The preview is the whole point of this route. You can click through several encodings and watch the sample change, without opening the file over and over. If nothing in the list gives readable text, the file may not be plain text at all, or it may use a writing system whose font isn’t installed (see the boxes and dots section below).

Which encoding to try first

Start with the language the file was written in, since many encodings are built for one writing system. Microsoft’s page lists encodings by writing system; this table points you to the right part of the list.

The file contains Look in the list for
Several languages in one file, or you don’t know the source Unicode entries, including UTF-8 (Microsoft groups these under Multilingual)
English or Western European text with accents Western European entries
Russian, Ukrainian or other Cyrillic text Cyrillic entries
Greek, Hebrew, Arabic, Thai or Vietnamese The entries named for that script or language
Japanese Japanese entries, such as Shift-JIS
Chinese Simplified Chinese (GBK or GB2312) or Traditional Chinese (Big5), depending on the source
Korean Korean entries

Turn the conversion prompt off again

With Confirm file format conversion on open turned on, Word shows the Convert File box every time you open a file that isn’t in a Word format. That’s useful while you sort out one problem file and tiring after that. Microsoft suggests switching it off again if you rarely need to choose an encoding: go back to File > Options > Advanced > General, clear the box, and select OK. Microsoft documents the setting on its confirm file format conversion page.

Pick an encoding when you save as plain text

The same File Conversion box appears when you save a document as a text file, so you can control how the next program reads it.

  1. Select File > Save As and choose where to save.
  2. In the Save as type list, select Plain Text, then select Save.
  3. If the Microsoft Office Word Compatibility Checker box appears, select Continue.
  4. In the File Conversion box, choose Windows (Default) for the system’s default encoding, MS-DOS for the MS-DOS standard, or Other encoding and a standard from the list.
  5. Check the Preview area, then select OK. You can resize the box to see more of the preview.

If you don’t choose an encoding, Microsoft says Word saves the file as Unicode. Unicode covers the most characters, so it’s the safest choice unless the program that will read the file needs something else.

Red text and Allow character substitution

When some characters can’t be stored in the encoding you picked, Word shows the message Text marked in red will not save correctly in the chosen encoding. You have two ways out: choose a different encoding that holds those characters, or select Allow character substitution. Substitution swaps each problem character for the closest equivalent, such as three periods for an ellipsis or straight quotes for curly quotes. A character with no equivalent is saved as a question mark.

Line breaks and line endings

Select Insert line breaks to break long lines, then use End lines with to pick a carriage return (CR), a line feed (LF), or both. Use this only when the program that reads the file expects lines of a fixed length or a particular line ending.

What a plain text file leaves behind

Plain text keeps characters only. Microsoft warns that text formatted in the Symbol font and field codes are removed when you save as encoded text, so keep the original .docx if you might need them. Saving in an encoding other than Unicode can also leave some characters unreadable later; Microsoft’s example involves Cyrillic and Hebrew text, which no single older encoding holds together.

Boxes or dots everywhere: a font problem, not an encoding one

If almost all the text looks the same, such as rows of boxes or dots, Microsoft explains that the font needed to show those characters may not be installed. Switching encodings won’t help in that case. Select some of the text and try another installed font that supports the script, or add a font for that language.

Microsoft also notes that when Word opens an encoded text file, it uses the fonts set on the Fonts tab of Web Options. If you open text in one script often, setting a suitable font there saves changing it each time.

A box can also stand for a character you meant to insert, not one Word misread. To add a specific symbol by its code, see typing Unicode characters in Windows.

When the garbled text isn’t Word’s doing

  • Garbled text in other programs too: when older programs on the PC show the wrong characters as well, the cause is usually a Windows setting rather than Word. Changing the system locale in Windows 11 deals with that case.
  • A document from another word processor: files like .wpd go through a converter, not the encoding box. Converting WordPerfect files to Word covers that route.
  • A Google Doc: download it as a .docx rather than as plain text, and it opens in Word without an encoding question. Converting a Google Doc to Word shows how.
  • A PDF: Word converts PDFs into editable documents through a separate process; converting a PDF to Word explains what to expect.

The full list of encodings, the writing systems they cover and Microsoft’s own wording are on its page about choosing text encoding when you open and save files.

Text encoding questions

Does Word save a text file as UTF-8 by default?

Microsoft’s wording is that Word saves as Unicode when you don’t choose an encoding. If another program specifically needs UTF-8, pick it yourself under Other encoding and check the preview.

Why did my curly quotes turn into straight quotes after I saved a .txt file?

Allow character substitution was selected and the chosen encoding couldn’t store curly quotes, so Word used the closest match. Save again with a Unicode encoding to keep them.

Can I fix a .docx that already has wrong characters in it?

Not with the encoding box, since Word’s own formats are Unicode and skip it. The wrong characters were usually pasted in or carried over from a converted file. Go back to the original source file and open it with the right encoding, then copy the text again.

Is there a File Conversion setting in Word for Mac?

Microsoft’s encoding and confirm-conversion pages give Windows steps only and don’t document a Mac route, so the steps above are for Word on Windows.

Does the Convert File box mean the file is damaged?

No. It only means the confirm option is on and the file isn’t in a Word format. Choose the file type that matches, such as Encoded Text for a .txt file.

Next step: once the file reads correctly, save a copy as a .docx so you don’t have to pick the encoding again, and turn the confirm option off if you don’t expect more text files.