ExcelAnalyzerClient-side file inspector

Excel used range checker

Verify that no file leaves your browser

Local check

Why Excel remembers rows you deleted

Each worksheet XML file can contain a dimension element, such as A1:XFD1048576. That rectangle is the used range Excel last recorded. It is supposed to hug the cells that matter. It often does not. Formatting, a stray character, a row that was cleared but not deleted, or an old list object can leave the dimension sitting on the last row of the grid. Excel then thinks the sheet is a million rows tall. Opening, saving, and copying become slower, and the file grows even though the visible table is short.

What this check compares

The checker reads the dimension ref and then walks the sheet once. A cell counts as real data when it has a value, a formula, or an inline string. Style-only cells are counted separately, because a formatted empty cell can extend the used range without holding a number you care about. The result shows the declared range, the last populated cell, and the wasted rows and columns between them. A phantom sheet, in the Audit score, is one whose declared range runs at least a thousand rows past the data and at least four times further than the real last row.

The repair steps live in the guide on what the dimension element is claiming.

How to confirm it in Excel

After this page names the sheet, press Ctrl+End on that sheet. If the active cell jumps far below the table, you are looking at the same problem. Select the first unused row through the last used row, right-click, and delete entire rows. Do the same for unused columns if the wasted-column count is large. Then save. Ctrl+End should land near the real table. Checking the file here again should show the dimension pulled back. Deleting values inside cells is not the same as deleting the rows. Clearing contents leaves the dimension where it was.

What this check will not do

It does not rewrite the file. It does not remove your formatting. On .xlsb workbooks the dimension lives in a binary sheet, so the page says the check is unavailable rather than pretending the sheet is empty. If a sheet part is larger than 15 MB uncompressed, cell-level scanning is skipped and the integrity note says so. Very hidden sheets are still scanned when their XML exists. Visibility is a separate question.

A small range can still be wrong

A dimension of A1:A200 on a sheet whose last value is in A20 is wasteful enough to warn, even if it is not large enough to move the Audit score as hard as a full-column phantom. Look at the wasted-row column rather than waiting for a red score. Printer settings and charts can also keep a distant cell alive. If deleting empty rows does not move Ctrl+End, inspect objects and defined print areas before you keep deleting.

Also check

Questions people ask before trusting the result

Will deleting rows destroy my formulas?

Deleting rows below the real table does not. Deleting rows inside the table does. Use the last real cell in the result as the boundary.

Why does the actual range start at A1?

The report shows the bounding box from the first column through the last populated row and column. It is a summary, not a list of every blank inside the table.

Do formatted blanks count as data?

They are reported as style-only cells. They do not move the last data cell, but Excel may still treat them as used. If the wasted count is small and style-only cells are large, clear formats past the table.

Can every sheet be fine except one?

Yes. The table is per sheet. One forgotten import sheet is enough to dominate save time.

Does this find the last cell of a chart?

It finds cells and row markers in the worksheet XML. A chart that refers to a distant range can keep that range alive even after cells look empty. Check the chart data source if deletion does not stick.

Last updated 5 November 2026. Initial release of this audit.