Microsoft Excel 2013 for Windows; modern Windows import alternative noted · CSV and text import · Intermediate
Import UTF-8 Semicolon CSV in Excel 2013 Without Broken Accents
Choose the source encoding and semicolon delimiter explicitly, preserve empty fields and codes, then verify accented text and decimal-comma amounts.
Official documentation reviewed; UTF-8 fixture and expected values independently checked; not tested in Excel · Editorial standards
Import the file through the text wizard, choose UTF-8 for File origin, and choose semicolon as the delimiter. These settings solve different problems: encoding controls characters such as é and ü; the delimiter controls where columns split. An encoding fix alone cannot turn a one-column import into the intended five columns.
Keep the original file untouched. If accents or identifiers have already been changed and the file saved over, reformatting cells cannot reliably reconstruct the original. Obtain another export instead of guessing replacements.
Use a fixture with visible failure cases
Download the UTF-8 semicolon text sample without BOM, or its otherwise identical UTF-8 copy with BOM. These are plain-text import files, deliberately supplied with .txt extensions so that opening them can invoke the text wizard. They contain no formulas, workbook formatting or saved cell types.
The unmodified content is:
Code;Name;City;Amount;Note
0007;José;München;1234,50;"Café; pickup"
0012;Zoë;Zürich;8,25;Ready
0020;François;;0,00;Pending
0031;Anaïs;São Paulo;12,75;Paid
The source contract is UTF-8, semicolon-delimited, a double-quote text qualifier, and comma decimals with no thousands separators. It has one header row and four data records. The blank City in the François record is intentional. The semicolon inside Café; pickup belongs inside a single Note field.
For a real .csv export, import it explicitly rather than double-clicking. If necessary, make a copy with a .txt extension to force the wizard; renaming an extension does not change the bytes, encoding or delimiter. Microsoft: importing text and CSV files.
Excel 2013 for Windows: legacy wizard
This is a documentation-based reconstruction of the legacy workflow, not a verified Excel 2013 build walkthrough. The intended route is Data > Get External Data > From Text, selecting the source file. This does not require the separate Power Query add-in. Microsoft’s Power Query history distinguishes the Excel 2010/2013 add-in from the later integrated import experience. The current wizard help documents the retained controls but no longer lists Excel 2013 in its Applies To section; see the verification limit below.
- In step 1, choose Delimited, start at row 1, and select the UTF-8 entry in File origin, commonly labelled 65001: Unicode (UTF-8). The wording and available choices must be checked in your build. If no UTF-8/65001 choice is offered, stop rather than selecting a different encoding or assuming compatibility. Also stop if the preview does not show José, München and São Paulo correctly; keep the original file and investigate the export encoding. Microsoft identifies code page 65001 as UTF-8.
- In step 2, select Semicolon and clear other delimiters. Choose
"as Text qualifier. Leave Treat consecutive delimiters as one unchecked so the empty City is retained. - In step 3, select Code in the preview and set it to Text. Set Name, City and Note to Text too. Leave Amount as General.
- Choose Advanced and specify decimal separator
,and thousands separator.for this fixture. There are no thousands marks here; the explicit setting prevents the decimal comma being interpreted using an unrelated system default. - Finish the wizard and import to A1 on a new blank worksheet. Do not overwrite a populated report. Widen columns to inspect the text.
Microsoft’s Text Import Wizard reference explains the encoding, delimiters, qualifiers, per-column types and Advanced controls. If your screen differs, stop and identify whether you opened the legacy wizard or the modern From Text/CSV preview; do not mix their steps.
Check shape before checking the total
The correct result occupies A1:E5. Confirm these values individually:
| Cell | Expected value | What it tests |
|---|---|---|
| A2 | Text 0007 |
Code was not converted to number 7 |
| B2 | José | Acute accent survived decoding |
| C2 | München | Umlaut survived decoding |
| E2 | Café; pickup | Quoted semicolon stayed in one field |
| C4 | Empty | Adjacent ;; preserved the missing City |
| D4 | Numeric zero | Empty City did not shift Amount left |
| C5 | São Paulo | Tilde and internal space survived |
If the header and every row are in column A, reimport with the correct delimiter. If E2 splits into two cells, check the text qualifier. If the François row shifts left, check the consecutive-delimiter option. Do not fix the four visible records manually and assume thousands of other rows will be correct.
Validate text and numeric meaning independently
Enter the checks outside A1:E5. They use English function names and comma argument separators; your locale may require localized names or semicolons. The import delimiter is a separate setting from a formula’s argument separator.
| Formula | Expected result | Purpose |
|---|---|---|
=EXACT(B2,"José") |
TRUE | Compare the imported name to the intended text |
=EXACT(E2,"Café; pickup") |
TRUE | Compare the whole quoted field |
=ISTEXT(A2) |
TRUE | Code is stored as text |
=EXACT(A2,"0007") |
TRUE | All four code characters remain |
=ISNUMBER(D2) |
TRUE | Amount is a numeric value |
=COUNT(D2:D5) |
4 | All four amounts, including zero, are numeric |
=SUM(D2:D5) |
1255.50 | Amounts reconcile to the source |
The numeric amounts are 1234.50, 8.25, 0 and 12.75; their display may use commas or periods according to your regional formatting. Inspecting a total alone is insufficient: SUM ignores text amounts, and a mixed numeric/text column can produce a plausible partial total. Use both the count and the row checks. See EXACT, ISTEXT, ISNUMBER, COUNT and SUM.
For longer identifiers, see import long identifiers as text in Excel 2013. Do not convert identifiers to numbers and then try to repair them with a display format. Microsoft’s leading-zero and large-number guidance explains why text type must be chosen before destructive conversion.
Why a BOM is helpful but not a complete import policy
A UTF-8 BOM is an encoding signature at the start of a file. Microsoft’s UTF-8 CSV guidance says a UTF-8 CSV with BOM can be opened normally; otherwise it recommends explicit import. That does not establish your semicolon, decimal or identifier choices. Both supplied files should yield the same five-column result when imported with the explicit settings above.
Use the BOM copy as a second controlled import on a separate sheet. Repeat the shape, accent and total checks. If the no-BOM version fails but the BOM version succeeds, record that difference and the exact Excel build instead of declaring every UTF-8 file safe to double-click. Never add a second BOM to an existing file or change the exporter without retaining the original.
When choosing UTF-8 does not restore the text
- The preview already shows
José: verify the chosen encoding against the exporter’s specification. If the actual source text was previously misdecoded and then saved, selecting UTF-8 again may faithfully read already-corrupted text. - The source itself contains
?instead of an accent: a lossy earlier export may have discarded the character. Re-export from the original system; a font change cannot recover missing information. - Names look right but amounts remain text: check the decimal setting and COUNT result. Do not strip punctuation from all fields, especially names and codes.
- Some records have extra columns: inspect their quoting in the original text. Do not globally delete semicolons, because the delimiter and the semicolon inside a quoted note have different roles.
- The source encoding is unknown: ask the exporter or inspect its documented settings. “Contains European names” does not establish that a file is UTF-8.
Once verified, save the working workbook as .xlsx while retaining the original text file. Do not overwrite the raw export with another CSV just to save formatting. For untrusted real exports, avoid evaluating unexpected formula-like fields; this synthetic sample contains none.
Modern Excel for Windows
In Microsoft 365, Data > From Text/CSV opens the modern import preview. Choose 65001: Unicode (UTF-8) and Semicolon, then use Transform Data to inspect column types before loading. If an automatic Changed Type step has converted Code to numbers, remove that step first and assign Code as Text from the original values. A later Text conversion cannot restore a lost leading zero. Set Amount’s numeric conversion locale to match its comma-decimal source, then repeat the checks above. Microsoft: Power Query data types and locale.
Alternatively, enable From Text (Legacy) under File > Options > Data, then open Data > Get Data > Legacy Wizards > From Text (Legacy). These modern enablement controls are not the Excel 2013 route. Mac and web import menus are outside this walkthrough.
Verification limits
Not tested in Excel, including Excel 2013. References were checked October 16, 2026. The legacy Windows procedure uses the documented text-wizard controls; current Microsoft Support pages do not certify an exact Excel 2013 build. UTF-8 bytes, BOM variants, five-column parsing, empty fields, text checks and the numeric total were independently checked outside Excel. No native screenshots, import execution, or earliest-version claim is made.