How to Split an Excel File into Multiple Files (by Sheet, Rows or Column Value)
One giant workbook is hard to send, import or share with the right people. Here is how to split it by sheet, by row count, by the value in a column or by date — keeping headers and formatting intact.
Updated October 22, 20265 min read
Big spreadsheets tend to grow into one workbook that does everything: every region's sales, every month's orders, every customer in one sheet. Then someone needs only the North region, an import tool rejects files over 10,000 rows, or each sales rep should see only their own accounts. Splitting the workbook is the answer — and in Excel itself it means filters, copying, pasting and saving, over and over. Done automatically, it takes seconds.
Four ways to split a workbook
| Split by | You get | Typical use |
|---|---|---|
| Sheet | One file per worksheet | A workbook with a tab per month, team or client |
| Row count | Files of N rows each, header repeated | Import limits, email size limits, batch processing |
| Column value | One file per distinct value in a column | One file per region, rep, store or category |
| Date | One file per date range | Monthly or quarterly files from one log |
Splitting a workbook, step by step
Add the workbook
Open the Excel splitter and add the XLSX file. Its sheets and columns are read so you can choose how to split.
Choose the split method
By sheet, by row count, by the value in a column, or by date range.
Set the details
For row counts, choose the number of rows per file — the header row is repeated in every part. For column values, choose the column whose values define the files.
Preview and download
Check the list of parts it will create, then download them together as one ZIP.
Split by sheet, row count, column value or date range; column widths, number formats and conditional formatting are kept, and every part arrives in one ZIP. Free: one workbook up to 25MB, up to 50 parts. Runs on your device.
Split an Excel fileWhat survives the split
A useful split keeps each part looking like the original: column widths, number and date formats, and conditional formatting all carry across, so a currency column still shows currency and a highlighted overdue row is still highlighted. Each part is a real XLSX file that opens in Excel, Google Sheets and LibreOffice.
Formulas that refer to other sheets or to rows that end up in a different part cannot keep working, because the data they point to is no longer in the same file. If the parts are meant for reading rather than calculating, that does not matter; if they are, convert those formulas to values first.
Splitting by column value: worked example
An order export has 40,000 rows with a "Region" column containing North, South, East and West. Splitting by that column produces four workbooks — orders-North.xlsx, orders-South.xlsx and so on — each with the header row and only that region's orders. Send each regional manager their own file, and nobody sees another region's customers.
This is also the cleanest way to share data under privacy rules: rather than trusting people not to scroll past their own rows, they simply never receive the others.
Doing it in Excel itself
Without a tool, splitting by value means: apply a filter, select one value, copy the visible rows, paste into a new workbook, save, and repeat for every value. For four regions it is tedious; for 200 sales reps it is an afternoon, and errors creep in. VBA macros can automate it, but they need writing, testing and macro-enabled files that many organisations block. An automatic splitter does the same in one step without macros.
Choosing a row count
- For import tools: use the importer's stated limit minus a small margin — 9,500 for a 10,000-row limit.
- For email: aim for files a few MB each; for plain data, roughly 50,000–100,000 rows.
- For people: a few thousand rows is what anyone will actually read.
After splitting
If a system needs CSV rather than XLSX, convert the parts — see Excel to CSV and JSON. If you need a read-only copy for someone without Excel, print each part to PDF on one page. And when the separate files come back with updates, the Excel merger combines them into one workbook again.
The workbook is processed on your own device, so customer lists, payroll and financial data never pass through a third-party server.
Splitting by date
Logs, transactions and orders usually have a date column. Splitting by date range turns one long history into monthly or quarterly files — orders-2026-01.xlsx, orders-2026-02.xlsx — which is what accountants, auditors and archives typically want. Make sure the date column contains real dates rather than text that looks like dates; text dates sort and group unpredictably. If the dates came from a CSV, check them first — see opening CSV files in Excel correctly.
When the file is shared, a date split also limits exposure: someone who needs last quarter's data receives last quarter, not the full five-year history.
CSV output and naming
Some systems import CSV rather than XLSX. On Pro, the parts can be written as CSV files directly, and both the file names and the ZIP name can follow a pattern you set — customers-{value}.xlsx, for example — so the files arrive ready for the system or the people they are meant for. On Free, the parts are XLSX with automatic names; converting a part to CSV afterwards is one step with the Excel to CSV converter.
Frequently asked questions
How do I split an Excel file into multiple files?
Add the workbook to the Excel splitter, choose to split by sheet, row count, column value or date, preview the parts and download them as one ZIP. Free handles one workbook up to 25MB, into up to 50 parts.
Can I split an Excel sheet by the value in a column?
Yes. Choose the column, and the splitter creates one file per distinct value — one per region, rep or store — each with the header row and only its own rows.
Is the header row kept in every file?
Yes. When splitting by rows or values, the header row is repeated at the top of every part.
Will formatting be kept?
Column widths, number formats and conditional formatting are kept. Formulas that point to data in a different part cannot keep working, so convert them to values first if the parts need them.