Skip to content
TroubleshootingVerified Bench 2026

Word Breaks Accented Characters | Fast Encoding Fix

You wrote café and naïve in Word, pasted into your CMS, and got café and ’ instead. The text wasn't corrupted by damage — it was encoded once by Word and decoded again by your website, and the two disagreed.

·5 min read·Last verified Sep 2026
Windows 11 24H2 & macOS 15.1

Why this happens

Word's paste path can carry text tagged with Windows-1252, an older single-byte encoding where é is one byte (0xE9). Almost every website runs on UTF-8, where é is two bytes (0xC3 0xA9). When your CMS receives that single Windows-1252 byte but interprets it as UTF-8, it reads garbage — producing é instead of é.

The same mismatch explains why Word's curly "smart quotes" turn into ’ or “ online — those characters sit outside the basic range some paste paths default to, so punctuation often mangles worse than accented letters.

The three fixes, in order of reliability

1. Paste as plain text. Ctrl+Shift+V (Windows) or Cmd+Shift+V (Mac) strips Word's encoding metadata entirely, so your website treats the text as fresh UTF-8 input. Formatting is lost and needs re-applying — a fair trade for text that isn't corrupted.

2. Disable smart punctuation before writing. File → Options → Proofing → AutoCorrect Options → uncheck "Straight quotes with smart quotes." This stops curly-quote substitution at the source (see our curly vs straight quotes guide).

3. Find-and-replace already-published damage. The corruption is a fixed, repeatable mapping — café always becomes café under this exact mismatch. A site-wide find-and-replace for the handful of sequences your content actually uses cleans existing pages without retyping anything. You can also sanitize text using our Text Cleaner tool.

Copy clean replacements:

Why some platforms are worse than others

Modern JS-framework sites and WordPress's block editor generally normalize pasted text to UTF-8 correctly. The mangling shows up more when text passes through a step that assumes Windows-1252 — an older backend without explicit UTF-8 declarations, a database column collation set to latin1 instead of utf8mb4, or a CSV round-trip (see our Excel special characters guide for that version). If the mangling happens on every paste rather than occasionally, the fix belongs in the backend configuration, not in how you paste.

Prevention for ongoing content work

If you write in Word regularly for the web, default to smart punctuation off and Ctrl+Shift+V for every paste into your CMS. One extra keystroke eliminates the failure mode, rather than fixing it after publication. For why encodings collide like this at all, see our deep dives on Unicode explained and UTF-8 vs UTF-16 vs ASCII.

Share:
JV

Dr. Julian Vance

Lead Editor

Lead Systems Architect & Digital Typographer at HowToTypeThis. Every keystroke, Alt code, hex sequence, and encoding behavior on this page was tested and verified on physical Windows 11 (24H2) and macOS 15.1 (Sequoia) hardware.

Looking for a different glyph or accent?Symbol Picker ↗

Last verified Sep 2026 · Windows 11 24H2, macOS 15.1 · How we test

View all 32 guides
Tray
Copied to clipboard!
Ready to paste anywhere