CSV vs CSV UTF-8 in Excel: Which Format to Save

CSV vs CSV UTF-8 is a choice you meet every time you save a CSV file from Excel's Save As list. Both create the same kind of plain text file with commas between values. The only difference is the character encoding, the rule that turns letters into bytes, and it decides whether names like Zoë, Łukasz or 李伟 survive the trip. This guide explains the difference in plain words, shows the same data saved both ways and tells you which to choose.

CSV vs CSV UTF-8 at a glance#

The two formats are identical except for how they store characters, and CSV UTF-8 is the safer choice for most data.

Both files open in Excel, Google Sheets, databases and most online tools. The difference shows up only when your data contains characters outside basic English.

What is character encoding?#

A computer stores text as numbers, and an encoding is the table that says which number means which letter. Two programs must use the same table, or the letters come out wrong.

• Regional encodings are older tables with room for about 256 characters. Windows-1252, used on English and Western European Windows, has plain English plus letters like é, ü and ñ. It has no room for Polish Ł, Greek, Arabic or Chinese.
• UTF-8 is a Unicode encoding. It covers every writing system in use, plus symbols like € and ✓. It is the standard encoding on the web.

When you choose plain CSV (Comma delimited), Excel uses your computer's regional encoding. When you choose CSV UTF-8 (Comma delimited), it uses UTF-8.

The same data saved both ways#

Here is a short customer list in Excel, saved on an English-language Windows computer:

Saved as CSV (Comma delimited)#

Windows-1252 has ë, ü, é, ú, ñ and ó, so the first two rows are fine. It has no Ł and no Chinese characters, so Excel replaces them with question marks:

Customer,City
Zoë Müller,Berlin
José Núñez,Madrid
?ukasz Wójcik,Kraków
??,??

The lost letters are gone for good. Opening the file again won't bring them back.

Saved as CSV UTF-8 (Comma delimited)#

Every character is stored, and the file reopens exactly as it was saved:

Customer,City
Zoë Müller,Berlin
José Núñez,Madrid
Łukasz Wójcik,Kraków
李伟,北京

Saving is only half the story. If a program reads that UTF-8 file as Windows-1252, José Núñez appears as José Núñez and Zoë as Zoë. The data is fine; it was just read with the wrong table. The fix is to open it with the right encoding, which the guide to importing a CSV file into Excel explains step by step.

When to use CSV UTF-8#

Choose CSV UTF-8 in almost every case. It is the right choice when:

• The data has names, addresses or product descriptions that could include any accent or alphabet.
• You are uploading to a website, online store, CRM, email tool or database. Most modern systems expect UTF-8.
• Someone in another country, or on a Mac or Linux computer, will open the file.
• You aren't sure. UTF-8 loses nothing.

Excel adds a byte order mark (BOM) to the start of a CSV UTF-8 file: three invisible bytes that tell programs "this is UTF-8". It helps Excel open the file correctly next time. A few older or strict programs don't expect it and show it as  in front of the first header, or reject the first column name. If that happens, check the receiving system's upload instructions.

When to use plain CSV#

Plain CSV (Comma delimited) still has a place. Use it when:

• An older system's instructions say it needs plain CSV, ANSI or Windows-1252 files.
• A program shows  at the start of your UTF-8 file and the data is plain English anyway.
• Your data has only basic English letters, numbers and punctuation. Then both formats produce the same bytes, apart from the BOM.

If the data has characters outside your regional encoding, plain CSV will lose them without a clear warning. Check the file before you send it.

How to save as CSV UTF-8 in Excel#

Saving as CSV UTF-8 takes four steps:

1. Click the sheet you want to save. A CSV holds one sheet only.
2. Go to File > Save As (Windows: F12; Mac: ⌘+Shift+S).
3. In Save as type (Mac: File Format), choose CSV UTF-8 (Comma delimited) (*.csv).
4. Click Save. If Excel warns that some features will be lost, confirm that you want to keep the CSV format.

Excel warns because CSV keeps only values. Keep an .xlsx copy as your working file. The comparison of CSV vs XLSX explains what each format keeps, and XLSX vs XLSM covers files with macros.

How to open a CSV UTF-8 file correctly#

Saving correctly is half the job; the other half is opening the file with the right encoding. Double-clicking a CSV lets Excel guess. With a byte order mark, it usually guesses right. Without one, it may fall back to your regional encoding and show é instead of é.

To control the encoding yourself:

1. Open a blank workbook and go to Data > From Text/CSV.
2. Select the file and click Import.
3. In the preview, set File Origin to 65001: Unicode (UTF-8) and check that the names look right.
4. Click Load, or Transform Data to set column types first.

This also lets you keep ID and phone columns as text, so leading zeros survive.

Common CSV encoding mistakes#

These problems come up again and again with CSV files:

1. Saving names with plain CSV. Letters outside the regional encoding become ? and can't be recovered. Use CSV UTF-8.
2. Double-clicking a UTF-8 file without a BOM. Excel may guess the wrong encoding. Use Data > From Text/CSV and choose 65001: Unicode (UTF-8) in File Origin.
3. Trying to "fix" é by hand. Reimport the file with the right encoding instead.
4. Expecting leading zeros to survive. Both formats store 00123 as text, but Excel turns it into 123 on opening. See how to keep leading zeros in Excel.
5. Ignoring hidden characters. Imported text can carry extra spaces or invisible characters, so lookups fail. The guide to why TRIM is not working shows how to remove them.

To clean up a header that picked up the BOM as Customer, remove the first three characters:

=RIGHT(A1,LEN(A1)-3)

LEN counts 11 characters, so RIGHT keeps the last 8: Customer.

Our recommendation#

Save as CSV UTF-8 by default. It keeps every character, it matches what most modern systems expect, and it costs you nothing in plain-English data. Switch to plain CSV (Comma delimited) only when a receiving system asks for it, or when it can't handle the byte order mark.

After any CSV import, check the text: look for odd symbols, extra spaces and numbers stored as text. Those are cleaning skills, and they are worth practising.

Practise it#

CSV files often arrive with messy text: stray spaces, invisible characters and numbers stored as text. Practise cleaning messy text in a live spreadsheet, then practise TEXT and VALUE to turn imported text back into numbers. Each step is checked instantly. Or try the exercise on this page and strip the BOM from a header.