Change UTF-8 in Outlook: Outgoing Mail and Files

Outlook uses the character encoding selected in File → Options → Advanced → International Options for outgoing messages, and for accented letters, currency symbols, and CJK characters to round-trip correctly that setting must be Unicode (UTF-8). Setting the mail client alone is not always enough: signature .htm files, contact .csv exports, and template files saved on disk also have to be valid UTF-8 bytes, because Outlook reads those files as text using the encoding declared inside or inferred from the file's bytes. When the file is actually Windows-1252 or UTF-16 but is labelled as UTF-8, Outlook shows replacement diamonds, scrambled accents, and CJK-looking squares in place of the intended characters. The same problem appears when a forwarded .msg file is saved using a legacy code page. The reliable fix is to make Outlook itself send in UTF-8 and to make sure every file you feed into Outlook is real, BOM-free UTF-8 — and that file side of the workflow is exactly what the UTF-8 Converter is built for.

Why Outlook Shows Garbled Characters

Outlook treats incoming and outgoing messages as a stream of bytes and a label that says which encoding those bytes follow. When the label and the bytes disagree, the characters that look identical in plain ASCII (A–Z, 0–9, basic punctuation) survive the mismatch, while everything that uses byte values above 127 — accents, the euro sign, curly quotes, emoji, CJK ideographs — gets remapped into whatever the wrong encoding thinks those bytes mean. The result is the classic mojibake pattern: José becomes José, € turns into ⬬, and 你好 can render as ä½ å¥½.

Two situations cause this in Outlook. The first is the mail client's own encoding choice. If Outlook is set to a Western code page such as Windows-1252 and you compose a message with non-ASCII characters, Outlook encodes those characters using the code page rather than UTF-8. Recipients on systems that expect UTF-8 see corruption even though your Outbox looks fine. The second situation is when Outlook imports a file — a signature, a contact .csv, or a stationery template — and the file itself is in a different encoding than Outlook assumed. Outlook cannot fix the file's bytes; it can only choose which decoder to apply, and picking the wrong one scrambles names and symbols.

Set UTF-8 as the Outgoing Mail Encoding in Outlook

For Unicode characters to round-trip through Outlook correctly, the outgoing mail encoder must be set to Unicode (UTF-8). The location of that setting has shifted across versions, but the option itself is the same. In desktop Outlook for Windows the path is:

1. Open File → Options.
2. Choose the Advanced tab. (In Outlook 2007 the equivalent control is on the Mail Format tab.)
3. Scroll to International Options and click it.
4. Open the Preferred encoding for outgoing messages dropdown.
5. Select Unicode (UTF-8) and click OK.

Restart Outlook so the new encoding takes effect on queued drafts. The same panel exposes a Preferred encoding for incoming messages dropdown; setting that to Unicode (UTF-8) helps Outlook pick a UTF-8 decoder when an inbound message omits an explicit charset, which is common with corporate relay servers. Microsoft documents the equivalent path for Outlook on the web (Settings → Mail → Compose and reply), where new messages are encoded as UTF-8 by design and the user does not pick a code page.

When the File Itself Needs to Be UTF-8

Setting Outlook's outgoing mail encoder to UTF-8 covers messages you compose and send. It does not cover files you load into Outlook, and these are the files that most often cause visible garbling:

• Signature files — the rich-text .htm signature and a plain-text .txt signature are both read from disk, so their bytes must be real UTF-8 if Outlook is to render curly quotes, accents, or CJK names inside the signature.
• Contact .csv exports — Outlook can import .csv files whose bytes were saved as Windows-1252 or UTF-16 by the exporting program; when those bytes are interpreted as UTF-8, names with accents, Asian characters, or symbols break at the import step.
• Stationery templates — custom .htm stationery is loaded as text by Outlook and rendered with the encoding the file declares.
• Quick Parts and AutoText .htt files — Outlook reads these from disk and uses the same text-decoding rules as stationery.

For each of these, the fix is to convert the file on disk to genuine UTF-8 bytes before Outlook opens it, rather than telling Outlook to reinterpret the wrong bytes.

Convert a File to UTF-8 for Outlook

Use the UTF-8 Converter to turn a signature file, contact export, or stationery template into a BOM-free, validated UTF-8 file. The conversion happens entirely in your browser; the source bytes are not uploaded.

1. Identify the file's actual source encoding. Look at the producing application or any reliable metadata — a .csv exported from an older Windows tool is almost always Windows-1252, a file saved from Notepad's "Unicode" option is UTF-16LE, and a file saved with "UTF-8 with BOM" is UTF-8. Do not guess from how the characters look, because the same byte sequences can mean different things under different legacy encodings.
2. Open the UTF-8 Converter and select the matching source encoding. The converter supports UTF-8 (with validation), UTF-16 little-endian, UTF-16 big-endian and Windows-1252.
3. Choose the file from disk. Files up to 10 MB are accepted; anything larger is rejected before reading, so the browser never accidentally loads a multi-gigabyte log.
4. Convert and review the preview. Read the decoded text and check names, punctuation, currency symbols and any non-ASCII line that matters. If UTF-8 is the source mode and the file contains invalid continuation bytes, truncated sequences, or forbidden encodings, the conversion fails loudly rather than silently emitting replacement characters.
5. Download the result. The downloaded file uses the original filename with -utf8 appended and a plain-text UTF-8 media type. No BOM is added; a recognised source BOM is consumed by the standards-based decoder.
6. Test the file in Outlook. Point Outlook at the new signature or import the converted .csv before deleting the original. The job identity on the converter prevents a slower old file read from replacing a newer choice, and obsolete object URLs are revoked once you change the selection.

The converter cannot detect your encoding for you, so if the source is genuinely unknown the workflow above has to start with finding the answer in the producing application or in any saved metadata — otherwise the conversion is just an arbitrary re-labelling of the bytes.

Source Encodings the Converter Handles

Each mode in the converter has a specific job. The table below summarises what each one is for and which Outlook problems it tends to fix.

Source mode	Typical Outlook scenario	Key behaviour
UTF-8	Files saved from modern editors, web exports, or another UTF-8 Outlook signature.	Validated with fatal error handling so malformed sequences fail rather than being silently replaced. A leading UTF-8 BOM is consumed.
UTF-16 little-endian (UTF-16LE)	Files saved by Notepad's "Unicode" option and many Windows export tools.	Decoded as low-byte first; a matching BOM is recognised and removed.
UTF-16 big-endian (UTF-16BE)	Files exported from some Unix, mainframe, or older Mac tools.	Decoded as high-byte first; picking LE instead of BE makes the letters come out as nonsense while bytes stay readable.
Windows-1252	Legacy Western .csv exports and .htm signatures copied out of older Word documents.	The browser's standards-defined Windows-1252 decoder supplies the printable punctuation and symbol mappings for bytes 80–9F, including the euro sign and curly quotes.

Outlook Desktop vs Outlook on the Web Encoding Controls

The two Outlook clients expose encoding very differently, and the converter only matters on the desktop side. The table below maps where each control actually lives.

Aspect	Outlook desktop (Windows)	Outlook on the web
Outgoing message encoding	User-selectable; set to Unicode (UTF-8) in International Options.	Always UTF-8 for message bodies; no user control.
Incoming message decoding override	User-selectable dropdown for ambiguous messages.	Driven by the sender's declared charset.
Signature file encoding	Read from disk; must be real UTF-8 if non-ASCII is used.	Not applicable — signatures are configured in Settings.
Contact .csv import encoding	Inferred from declared charset; UTF-8 imports render correctly.	Not applicable — contacts are managed via People.

Testing the Converted File in Outlook

After downloading the converted file, point Outlook at it before deleting the original. For a signature, open File → Options → Mail → Signatures, click New, browse to the new file, and confirm the preview pane shows every accented name, every curly quote, and every symbol you care about. For a .csv import, use File → Open & Export → Import/Export → Import from another program or file → Comma-Separated Values, choose the converted file, and map each column. Spot-check imported names and notes rather than trusting the import to finish without warnings.

Watch the source and output byte counts the converter reports. Different encodings use different byte lengths for the same characters, so a correct UTF-8 conversion very often changes file size — the Windows-1252 euro byte 80, for example, becomes the three UTF-8 bytes E2 82 AC. A different count is expected and does not by itself indicate data loss; what matters is whether the preview's characters match what the file is supposed to contain.

Retain the original file until the complete workflow is verified. If the converter's preview looked fine but Outlook still shows corruption, the source encoding was probably identified incorrectly the first time; re-export from the producing application with a known encoding label or metadata, then run the converter again with that mode. Files containing a mix of encodings inside one document cannot be repaired reliably by a single decoder — split the file by source first if that situation applies.

If the file you are feeding into Outlook is a plain text snippet whose source encoding is already UTF-8 but the file has become corrupted, the change UTF-8 in Notepad when the file looks broken workflow walks through the same validation approach from the editor side.

For a deeper look, see UTF-8 Browser Tools: Privacy and Round-Trip Verification.