[Excel VBA] How to Correctly Read CSV Files by Specifying Character Encoding (UTF-8, etc.)

2025年12月30日

Have you ever experienced garbled text (Mojibake) when opening a CSV file in Excel?

Especially when you directly open a CSV file saved in UTF-8 format in a Japanese environment Excel, full-width characters and symbols often get garbled.

In this article, I will introduce how to correctly read text files by specifying the character encoding using Excel VBA.

Target: UTF-8 Format CSV File

For example, let’s assume the following CSV file (e.g., product_utf8.csv):

Product Name	Price	Remarks
Apple	100	Grown in Aomori
Banana	80	Grown in Philippines

If you open this directly in Excel, the text may be displayed as “??” or “Ã¦”.

The cause is that Excel is not opening the file with the correct character encoding (UTF-8).

VBA Code: Reading CSV in UTF-8

Sub ImportCSVwithUTF8()
    Dim csvPath As String
    csvPath = ThisWorkbook.Path & "\product_utf8.csv"
    
    Workbooks.OpenText _
        Filename:=csvPath, _
        Origin:=65001, _ ' UTF-8 (Code Page)
        DataType:=xlDelimited, _
        Comma:=True
End Sub

Explanation of Arguments

Parameter	Explanation
Filename	The path of the CSV file to read.
Origin:=65001	Specifies the UTF-8 code page (65001).
DataType	Specifies the delimiter format (`xlDelimited`).
Comma:=True	Reads the file as comma-separated.

Reference: List of Code Pages (Partial)

Code Page	Corresponding Character Encoding
932	Shift-JIS (Japanese)
65001	UTF-8
1200	UTF-16 (Unicode)

Image After Execution (Correctly Displayed)

Product Name	Price	Remarks
Apple	100	Grown in Aomori
Banana	80	Grown in Philippines

Important Notes

The specification of Origin is very important. If omitted, Excel tries to open it with the OS default code (e.g., Shift-JIS), which causes garbled text.
UTF-8 CSVs are common in files output from web apps and systems.
Workbooks.OpenText is effective not only for text files but also for CSVs.

Application Points

For semicolon-separated or tab-separated files, you can use Semicolon:=True or Tab:=True together.
Even UTF-8 with BOM (Byte Order Mark) can basically be read with this method.