CSV Encoding Explained: UTF-8 vs Windows-1252 vs ISO-8859-1 (2026)

1. Why Encoding Matters for CSV

Every CSV is a sequence of bytes. The same byte value can mean different characters in different encodings. For example, the bytes C3 A4 in UTF-8 are the character ä; in ISO-8859-1, E4 is ä. If your app or Excel assumes the wrong encoding, you see or ä instead of ä. For more on CSV structure, see What is a CSV File?.

2. UTF-8: The Modern Standard

UTF-8 can represent every character in Unicode. It uses 1 byte for ASCII (a–z, 0–9) and 2–4 bytes for accented letters, umlauts, and symbols. UTF-8 is the default for the web, modern APIs, and most databases. For CSV, use UTF-8 whenever possible and save with a BOM (Byte Order Mark: EF BB BF) if you need Excel on Windows to open it correctly without asking.

name,city
Müller,Zürich
François,Paris
Niño,Madrid

3. Windows-1252 (CP1252)

Windows-1252 (Code Page 1252) is the default in older Windows apps and Excel in many locales. It’s a superset of ISO-8859-1 with extra characters in the 0x80–0x9F range (e.g. smart quotes, Euro €). CSV exported from legacy Windows systems is often Windows-1252. If you open such a file as UTF-8, characters like ä, ö, ü can appear as two-character sequences (e.g. ä).

4. ISO-8859-1 (Latin-1)

ISO-8859-1 covers Western European languages with one byte per character (256 code points). It’s common in older European systems. It’s very similar to Windows-1252; the main difference is the 0x80–0x9F range. Many tools treat “Latin-1” and “CP1252” interchangeably for CSV, but strictly they’re not identical.

5. How to Detect and Fix Encoding

Detect: If you see patterns like ä (ä), ö (ö), é (é), the file is likely UTF-8 being read as Windows-1252 or Latin-1. The reverse (wrong characters when you expect Latin-1) means the file is Windows-1252/Latin-1 but opened as UTF-8. Use a validator or a small script to try both. neatcsv’s CSV Validator can help spot encoding issues; Clean CSV can normalize output to UTF-8.

Fix in Excel: Use “Data → From Text/CSV”, choose the file, then in the import wizard select the correct “File origin” (e.g. “65001: Unicode (UTF-8)” or “1252: Western European”) and load. Save As → CSV UTF-8 (Comma delimited) to export as UTF-8 with BOM. More in How to Open CSV in Excel.

6. Code Examples: Python & JavaScript

Reading CSV with the right encoding and writing UTF-8 keeps data consistent.

Python

# Read Windows-1252 CSV, write UTF-8
with open('input.csv', encoding='cp1252') as f:
    text = f.read()
with open('output.csv', 'w', encoding='utf-8-sig', newline='') as f:
    f.write(text)  # or use csv.reader/writer for structured data

JavaScript (browser / Node)

// File API: read as UTF-8 (default for text())
const text = await file.text();
// If you have bytes from a legacy system, use TextDecoder
const decoder = new TextDecoder('windows-1252');
const text = decoder.decode(uint8Array);

For full CSV parsing (headers, commas, quotes), use a library like Papa Parse (JS) or the built-in csv module (Python) and always pass the correct encoding when reading.

7. Best Practices

• Save new CSV as UTF-8, preferably with BOM if Excel on Windows will open it.
• Document the encoding in specs or filenames (e.g. export_utf8.csv) so downstream tools don’t guess.
• When importing, try UTF-8 first; if you see mojibake, try Windows-1252 or ISO-8859-1.
• Use a CSV validator or a small script to detect encoding (e.g. Python chardet) before processing.

For more on fixing broken CSV (columns, delimiters, dates), see 10 Common CSV Errors.