Keep only the columns you need from a huge CSV, without opening it in Excel
Wide exports are the other half of the big-CSV problem. One analyst on Reddit had a 997 MB file with 1.8 million rows and 228 columns, on Excel 2013 with no outside programs allowed, and needed four of them: A, GG, GM and GZ. A data engineer had a 75 GB vendor dump with about 400 columns and needed 38. Someone opening the public NPPES health-provider file (5.29 GB) put it plainly: “I really don’t need probably 85% of it.”
Cutting the columns is often all it takes to make the file fit: 4 columns out of 228 is under 2% of the width, so a file Excel can’t open becomes one it opens instantly.
MegaCSV does it in your browser: open the file, tick the columns you want, export.
Key takeaways
- Open the CSV in MegaCSV, choose Export, then Columns → Only the ones I choose… and tick the ones you need. A search box finds a column by name among hundreds.
- The CSV export copies each chosen value byte for byte, with the file’s own delimiter, so leading zeros, long IDs, quotes and dates arrive unchanged.
- Combine it with a filter to cut rows and columns in one export, and with Split if the result is still over Excel’s 1,048,576 rows.
- Your file is never uploaded, and there’s nothing to install.
Why deleting columns in Excel doesn’t work for big files
The usual advice is “open it and delete the columns you don’t need”. That fails twice on a big file:
- Excel can’t open all the rows. Anything past row 1,048,576 is dropped as the file opens (every Excel limit), so the trimmed file you save is missing data.
- Excel changes the values it does open. Leading zeros disappear, long numbers lose digits and dates are rewritten, and saving back to CSV makes the changes permanent (what Excel changes).
Power Query can remove columns without the row limit, on Windows, but it’s a query to build and you still need a way to get the result out. Command-line tools like cut work for simple files but break on commas inside quoted values.
How to keep only some columns with MegaCSV
- Open the file. Go to megacsv.com/app and drop the CSV on the page. The first rows appear straight away, whatever the size.
- Find the columns you need. Scroll right through the header, or note the names from the vendor’s spec.
- Optional: filter the rows. Choose Filter to keep, for example, one country or one year. Only matching rows are exported.
- Choose Export. Under Columns, pick Only the ones I choose…. Choose Select none, then type part of a name in Find a column and tick each one you need. The count shows how many you’ve picked, for example 4 of 228 columns.
- Pick the format. CSV copies each chosen field exactly as it was written. Excel (.xlsx) writes IDs, leading zeros and long numbers as text so Excel can’t change them.
- Export. The header row is cut to the same columns.
Columns are written in the order they appear in the file.
Frequently asked questions
How do I delete columns from a CSV that’s too big to open?
Open it in MegaCSV, choose Export → Columns → Only the ones I choose…, tick the columns to keep and export. The original file isn’t changed; you get a new file with just those columns.
Will removing columns change my data?
No. In a CSV export, each value you keep is copied byte for byte, including quotes, and joined with the file’s own delimiter.
Can I keep columns and filter rows at the same time?
Yes. Apply a filter first, then choose the columns in the export dialog. The export contains only matching rows and only the chosen columns.
Can I reorder the columns?
Not in MegaCSV: columns keep the order they have in the file. Reorder them in Excel after the export, when the file is small enough to open.