Splitting a file
Paste or open the CSV
Paste the data or use Open a file. Check the row and column counter, and set the separator if detection picked the wrong one.
Set rows per file
Type the number of data rows each part should hold. The default is 1,000. The summary shows how many files that makes and how many data rows there are in total.
Keep the header in every part
Repeat the header row in every file is on by default, so each part imports on its own with the right column names. Untick it for headerless data.
Download
Download all builds a zip of every part, named part-1, part-2 and so on with leading zeros so they sort correctly. Below it, a button for each part (up to the first 50) shows its row count.
Five leads into parts of two
A tiny list makes the behaviour easy to see. With Rows per file set to 2 and the header repeated, five data rows become three files: two full parts and a last part with the single remaining row.
The header counts as an extra line in each file, not as one of the two data rows. That matters when an import limit includes the header line: if a tool accepts at most 500 lines including headers, set the splitter to 499.
business,town,email Kiln & Co Pottery,Frome,hello@kilnandco.example Marlow Print Works,Bath,info@marlowprint.example Oak Lane Joinery,Frome, Tanner's Yard Cafe,Bath,hi@tannersyard.example Pip & Parsnip Deli,Wells,shop@pipparsnip.example
Files: 3 Data rows: 5 Rows per file: 2 part-1.csv header + Kiln & Co Pottery, Marlow Print Works part-2.csv header + Oak Lane Joinery, Tanner's Yard Cafe part-3.csv header + Pip & Parsnip Deli
Splitting well
Rows stay in the order they were pasted, so the first part always holds the top of the file. If you want each batch to be a fair mix, for example a spread of towns for each sales rep, sort or shuffle the file before splitting. If you want each part to be one region, sort by postcode first and pick a size that roughly matches.
De-duplicate before you split, not after. A duplicate pair that lands in two different parts cannot be caught by checking each part on its own, and one business ends up in two people's call sheets.
Before you upload the parts
- The header is repeated, unless the destination expects headerless files.
- The row count per part fits the destination's limit, allowing for the header line if it counts.
- Duplicates were removed from the whole file first.
- The parts open correctly in a spreadsheet; special characters look right.
Questions people ask
How do I split a CSV file into multiple files?
Paste or open the CSV, enter the number of rows each file should contain, keep the header option on, and press Download all to get a zip of numbered parts. Each part is a normal CSV with the same columns, so it opens in Excel or imports into a CRM on its own.
Does each split file keep the header row?
Yes, by default. Repeat the header row in every file copies the first row of your data to the top of each part. If your file has no header, untick the option so the first data row is not mistaken for column names and copied into every part.
Why can I only see buttons for 50 parts?
To keep the page quick, individual download buttons are shown for the first 50 parts only. Download all still includes every part in the zip, however many there are, so use that for big splits and open the zip on your computer.
Can I split by a column value instead of a row count?
Not with this tool: it splits by number of rows. To make one file per town or per region, open the data in the CSV viewer, search for the value, and download the matching rows, then repeat for each value you need.
Where your data goes
The parts and the zip are created by your browser from the data on this page. The CSV is never sent anywhere, and the zip library loads only when you press the download button.
Checked against the tool on