How to Fix UTF-8 CSV Encoding in Excel (BOM & Delimiters)
Few bugs waste more analyst time than exporting a perfectly valid UTF-8 .csv file from an API or database, double-clicking it in Microsoft Excel, and seeing Müller turn into Müller, São Paulo turn into São Paulo, or Japanese and Chinese text (東京, 深圳) collapse into unreadable symbols. Knowing how to fix UTF-8 CSV encoding in Excel—and why European Excel installations sometimes dump every column into Column A—comes down to three bytes at the start of the file and your operating system's regional delimiter setting.
Let's examine why desktop Excel misreads standard UTF-8 files on double-click, and walk through three reliable fixes depending on whether you are generating the CSV or opening an existing file.
Why Microsoft Excel Garbles Valid UTF-8 CSV Files
Plain-text .csv files carry no internal metadata header declaring their character encoding. When a developer exports UTF-8 text from Python, Node.js, PostgreSQL, or a REST script without a Byte Order Mark, the file starts immediately with ASCII header bytes (69 64 2C 6E 61 6D 65 for id,name).
When you double-click that .csv file in Windows Explorer or macOS Finder, Microsoft Excel historically falls back to the system's legacy 8-bit code page (such as Windows-1252 in Western locales or Shift-JIS / GBK in East Asian locales) instead of guessing UTF-8. Because UTF-8 encodes ü as two bytes (C3 BC), Windows-1252 interprets C3 as à and BC as ¼—producing the classic mojibake string Müller.
Fix 1: Prepend a 3-Byte UTF-8 BOM (EF BB BF) Before Opening
The cleanest permanent way to fix UTF-8 CSV encoding in Excel so anyone on your team can simply double-click the file is prepending the UTF-8 Byte Order Mark (BOM): hexadecimal bytes EF BB BF (Unicode character \uFEFF) at byte offset 0.
When Excel detects EF BB BF at the beginning of a .csv file, it bypasses legacy code-page guessing and decodes every row as UTF-8 immediately:
- Zero Wizard Steps: Recipients on both Windows and macOS can double-click the
.csvattachment directly without navigating import menus. - One-Click Browser Fix: If you are converting JSON payloads, paste or drop your data into our client-side JSON to CSV converter with Include UTF-8 BOM checked (on by default). Your downloaded
.csvautomatically includes\uFEFFat byte0while keeping your data 100% inside local browser RAM (see our privacy policy). - In JavaScript / Node.js / Python:
- JavaScript Blob:
new Blob(["\uFEFF" + csvText], { type: "text/csv;charset=utf-8;" }) - Python
pandas:df.to_csv("export.csv", index=False, encoding="utf-8-sig")
- JavaScript Blob:
Note on CLI pipelines: If you are feeding a CSV into strict UNIX command-line utilities or legacy parsers that do not strip BOM headers, uncheck Include UTF-8 BOM so byte 0 starts directly with the first column name, or export a native .xlsx workbook via our JSON to XLS tool instead.
Fix 2: Import an Existing BOM-Less CSV via Excel's Data Ribbon
If someone already sent you a UTF-8 .csv file without a BOM and you cannot re-export the source, do not double-click the file (and never click Save after seeing garbled text, which overwrites the original bytes permanently). Instead, tell Excel explicitly to use 65001: Unicode (UTF-8):
- Open a blank workbook in Microsoft Excel.
- Go to the Data tab on the top ribbon and click From Text/CSV (in the Get & Transform Data group).
- Select your
.csvfile and click Import. - In the preview dialog, change File Origin at the top left to
65001: Unicode (UTF-8). - Verify the Delimiter dropdown (
Comma,Semicolon, orTab) so preview columns align properly. - If your CSV includes postal codes or SKU IDs with leading zeros (
00491), click Transform Data to set those columns to Text before clicking Load.
Fix 3: Solve the "Everything in Column A" Regional Delimiter Trap
Even after you fix UTF-8 CSV encoding in Excel, teams working across North America and Europe often hit a second snag: the text characters look crisp, yet every row sits unsplit inside Column A (id,name,city,amount).
Here is why: Excel ties its default CSV field separator to your OS decimal symbol:
| Regional Decimal Format | Typical Countries | Excel Default CSV Separator |
|---|---|---|
Period Decimal (12.50) |
US, UK, Canada (EN), Japan, China, Australia | Comma (,) |
Comma Decimal (12,50) |
Germany, France, Spain, Italy, Netherlands, Brazil | Semicolon (;) |
When sending CSV exports to colleagues in Germany, France, or Brazil, either select Semicolon (;) + UTF-8 BOM in our JSON to CSV converter, or export a genuine binary .xls / .xlsx file from JSON to XLS—or from .docx reports using our Word to Excel converter—because native Excel workbooks store cells structurally and ignore regional delimiter settings altogether.