Guides / CSV too large for Excel

Beyond the grid

Your CSV is too large for an Excel worksheet. It is not too large to inspect.

Excel worksheets stop at 1,048,576 rows. Preserve the source, inspect all rows in a file-backed viewer, and export only the useful result.

1,048,576maximum rows in a current Excel worksheet
50,000,000rows in csvlite’s observed 8 GB test file

First, do not split blindly

Splitting a CSV into worksheet-sized chunks makes global sorts, duplicate checks and aggregations easy to get wrong. Keep an untouched original. If Excel is the final destination, filter or aggregate the full file first and export a smaller result.

Your local options

  • csvlite: visual, read-only filtering, profiling and pivots without syntax.
  • DuckDB: powerful SQL against the CSV directly.
  • Power Query/Data Model: stays in the Microsoft ecosystem, but does not place every row in the worksheet grid.

Microsoft explicitly recommends checking source and destination counts and documents alternatives when data exceeds the grid. See Microsoft’s guidance.

Export the question, not the whole file

In csvlite, stack filters, choose the visible columns, and export the current view. Or group by categories and calculate counts, sums, averages, minima, maxima or distinct values. The exported CSV becomes a manageable spreadsheet while the original remains the source of truth.