Home › Guides › Fix CSV problems

Fix garbled characters, lost leading zeros and other CSV problems

Why CSV files show é instead of é, drop leading zeros, turn long numbers into 1.23E+15 or swap dates, and how to fix each one in Excel.

Updated October 2026

You converted an Excel file to CSV, opened it, and something is off: accents look like é, ZIP codes lost their zeros, or long numbers turned into 1.23457E+15. In almost every case the CSV is fine. The damage happens when Excel opens it. Here's what each problem means and how to fix it.

Accents and symbols look like é or ’

Your CSV is saved as UTF-8, the modern standard. When you double-click it, Excel may assume an older Windows encoding instead and read each byte as a separate letter. So é shows as é, and a curly apostrophe shows as ’.

Fix 1: add a BOM. Convert the file again on xlsxtocsv.com with Encoding set to UTF-8 with BOM. The BOM is three invisible bytes at the start of the file that tell Excel "this is UTF-8". Excel then opens it correctly with a double-click.

Fix 2: import instead of opening. In Excel, go to Data › From Text/CSV, choose the file, and set File Origin to 65001: Unicode (UTF-8). Check the preview, then click Load.

Leading zeros disappear (00501 becomes 501)

Open the CSV in Notepad (Windows) or TextEdit (Mac) and you'll see the zeros are there. Excel removes them because it treats anything that looks like a number as a number. That breaks ZIP codes, phone numbers, product codes and employee IDs.

Fix: tell Excel the column is text.

1. In a blank workbook, go to Data › From Text/CSV and pick your file.
2. Click Transform Data (not Load). The Power Query editor opens.
3. Click the header of the column with the zeros, then choose Data Type › Text. If Excel asks, choose Replace current.
4. Click Close & Load. The zeros stay.

In Microsoft 365 for Windows you can also stop the stripping for good. Go to File › Options › Data, find Automatic Data Conversion, and clear Remove leading zeros and convert to a number.

Long numbers turn into 1.23457E+15

Excel stores numbers with 15 digits of precision. A 16-digit card number or an 18-digit tracking number loses its last digits (they become zeros) and shows in scientific notation. The CSV still has every digit, as long as the original .xlsx did.

Import the file the same way as for leading zeros and set that column to Text. In Microsoft 365, you can also clear Keep first 15 digits of long numbers and display in scientific notation where needed under Automatic Data Conversion.

If the number was already rounded in the .xlsx file (it ends in zeros there), the missing digits can't be recovered from it. You'll need the original source data.

Dates change or swap day and month

A CSV stores dates as plain text, such as 03/04/2026. Is that March 4 or April 3? It depends on the computer that opens it. US settings read month first, most other countries day first.

Fix: convert with Dates set to ISO (2026-03-04). The year-month-day format means the same everywhere, and every database, CRM and spreadsheet reads it correctly.

Everything is squashed into one column

The separator in the file doesn't match what Excel expects. Excel in much of Europe expects semicolons, because the comma is the decimal mark there. Convert again with Separator set to Semicolon. Or use Data › From Text/CSV and pick the delimiter in the import window.

Rows break in the middle

A cell that contains a line break (Alt+Enter in Excel) is written inside quotes, which is correct CSV. Some importers still treat the line break as a new row. Before converting, remove line breaks in Excel. Press Ctrl+H, click in Find what and press Ctrl+J (it looks empty, but it now holds a line break), type a space in Replace with, and click Replace All.

Extra empty rows or commas at the end

Excel remembers cells that once had content or formatting, so the "used range" can stretch past your data. That shows up as rows of commas. Check Skip empty rows in the converter, or delete the empty rows and columns below and to the right of your data in Excel, save, and convert again.