A text format versus an application file format
CSV is plain text — rows and comma-separated values, nothing more, readable by literally any text editor and parseable by a few lines of code in any language (see the Complete Guide to CSV Files). Excel's native format (.xlsx) is a structured, zipped bundle of XML files representing a full spreadsheet application's feature set: multiple sheets, cell formatting (colors, fonts, borders), formulas with live recalculation, charts, and more. CSV can represent only the first layer — flat, unformatted tabular data — none of the rest.
What's lost converting Excel to CSV
Exporting an .xlsx file to CSV keeps the raw cell values but discards nearly everything else: cell formatting, multiple sheets (CSV supports exactly one flat table, not a workbook), and — critically — formulas become their last-calculated static value, not the formula itself. A cell showing =SUM(A1:A10) exports as whatever number that formula last evaluated to, with no way to recover the original formula from the CSV alone.
| Feature | Excel (.xlsx) | CSV |
|---|---|---|
| Multiple sheets | Yes | No — one flat table only |
| Formulas (live, recalculating) | Yes | No — exports as static last value |
| Cell formatting (colors, fonts) | Yes | No |
| Charts | Yes | No |
| Readable by any text editor | No (binary/zipped XML) | Yes |
| Parseable with minimal code | No | Yes |
Why CSV remains the interchange default anyway
Despite losing all of Excel's richer features, CSV's simplicity is exactly why it remains the universal interchange format between systems that don't share Excel's proprietary format — any database, script, or unrelated application can read and write CSV without needing Excel-specific libraries, while .xlsx requires dedicated parsing support for its more complex structure. When moving data between systems (not just within Excel), CSV is often the lowest-common-denominator choice precisely because it gives up everything Excel-specific.
Common mistakes
- Assuming a CSV export preserves formulas. It only keeps the last-calculated static value — the underlying formula logic is gone, not recoverable from the CSV file itself.
- Expecting multi-sheet Excel workbooks to export cleanly to one CSV. CSV has no concept of multiple sheets — an export typically covers only the active sheet, or requires separate CSV files per sheet.
- Losing number formatting on re-import. A cell formatted to show as a percentage or currency exports as the underlying raw number in CSV — the display formatting itself doesn't survive the round trip.
FAQ
Does exporting Excel to CSV keep the formulas?
No — only the last-calculated static value exports; the formula itself is lost and can't be recovered from the CSV alone.
Can a CSV file represent multiple Excel sheets?
No — CSV is a single flat table format; a multi-sheet workbook needs either multiple separate CSV files or a format that natively supports sheets, like .xlsx itself.
Why is CSV still preferred for moving data between different systems?
Its simplicity — any tool or language can parse it with minimal code, unlike .xlsx's more complex, application-specific structure requiring dedicated library support.
Inspect and convert structured data with the JSON Toolkit — entirely client-side.