Spreadsheets convert badly to PDF more often than any other format, and the reason is structural rather than a failing of any particular tool.
A worksheet has no pages. It is an unbounded grid. Paper has fixed dimensions. Converting one to the other means inventing page boundaries that the source document never defined — and if you do not decide where they fall, the conversion decides for you, usually badly.
The cut has to happen somewhere
A grid twenty columns wide will not fit across a portrait page. Something must give:
- Cut the columns and continue them on a later page — the default, and the source of the familiar complaint that a table arrived in unreadable fragments.
- Shrink everything until it fits, which can render the text too small to read.
- Turn the page sideways, which buys perhaps forty per cent more width.
- Print fewer columns, by deleting the ones the reader does not need from a copy of the workbook.
None of these is universally right, which is why the decision belongs to you and belongs in the workbook, before conversion.
Set these before converting
Excel’s own Save as PDF follows the sheet’s page setup: print area, fit-to-page scaling and print titles. Excel to PDF reads only the orientation from it, so four things in the workbook do almost all the work:
- What is in the cells, not the print area. The print area is not read: every cell with a value converts, including cells outside it — which is how a three-column table becomes several pages, most of them blank, because a stray value sits in column BQ. Delete anything that must not appear, from a copy of the workbook if you need to keep it.
- Orientation. Pages are landscape unless the sheet’s page setup says portrait, so set portrait only for a narrow sheet.
- Column widths, not scaling. Fit-to-page is not read and nothing is shrunk, so a wide sheet continues on further pages in bands of columns. Narrower columns mean fewer bands.
- A frozen header row. Rows to repeat at top are not read, but rows frozen with Freeze Panes repeat at the top of every page — the difference between a readable multi-page table and a wall of anonymous numbers.
Each sheet is drawn from what is stored in the file — cached values, number formats, column widths, fills and bold or italic text — so getting the workbook right is still most of the job.
Formulas become values, permanently
A PDF has no calculation engine. Every formula converts to the value it last produced.
This is usually what you want — a report should show results, not =SUM(B2:B47). But it has a sharp edge.
Spreadsheets store cached results alongside formulas, and the conversion reads those cached values. If the workbook was saved without recalculating — a real possibility with manual calculation enabled, or with external links that did not refresh — the PDF faithfully reproduces stale numbers.
There is no way for the conversion to detect this. The cached value is all there is.
Open the workbook, force a recalculation, save, then convert. For anything financial, that sequence is not optional.
Dates, currency and the formatting trap
A date in a spreadsheet is a number. 45000 with a date format attached displays as a date; the same number with a general format displays as 45000.
The format is the only expression of intent in the file, so a column of dates that someone pasted as values without applying a date format converts as a column of five-digit numbers — correctly, and uselessly.
The same applies to currency, percentages and decimal places. A cell showing 1.5 that is actually 1.4999999 converts as 1.5, because that is what the format says to display. That is right for a report and worth knowing when the numbers are checked against a system that holds full precision.
Excel to PDF has three exceptions for now. Any date format is written year-month-day, such as 2026-03-15, whatever pattern the sheet uses; a cell holding only a time comes out with a placeholder date in front, such as 1899-12-30 12:00; and Accounting format prints a stray asterisk, such as $* 1,234.50.
Fix formatting in the workbook. Nothing downstream can infer what was meant.
What does not survive
Be aware of what the conversion drops, since none of it is recoverable afterwards:
- Formulas — values only, as above
- Multiple sheets — each becomes its own set of pages, in workbook order, with nothing linking them
- Hidden rows and columns — not a safe way to keep data out. Rows hidden in Excel are left out, but a hidden column’s contents can still print, depending on how the workbook stores it, and rows and columns hidden in LibreOffice can print. Delete anything confidential rather than hiding it.
- Comments and notes — generally omitted
- Interactivity — filters, sorting, anything a reader could click
The last one is the reason to think about audience. A colleague who needs to sort and filter wants the workbook. A client who needs a fixed record wants the PDF. Sending the wrong one wastes everybody’s time.
Long tables across many pages
For a table running to dozens of pages, two things make it usable:
- Repeat the header row on every page by freezing it with Freeze Panes; Excel to PDF does not read print titles.
- Add page numbers to the converted file with Add page numbers, so a reader can refer to a specific page.
If the output is a reference document destined for print, Create PDF booklet can place two pages per sheet — halving the paper for a table people will flip through rather than read linearly.
Going back the other way
Sometimes you receive the PDF and need the numbers. PDF to Excel reconstructs a table by reading the positions of the text on the page.
Set expectations correctly: you get values arranged in rows and columns. Formulas, formatting and sheet structure are gone — they were discarded at conversion and no tool can recover them. For a clean, ruled table the reconstruction is good. For merged cells, nested headers or a table split across pages, check it against the original before trusting a total.
Before you send it
- Check the page count. Far more than expected usually means a stray value far to the right of the table, which adds bands of empty columns; delete it and convert again.
- Look at the right edge of the first page to see whether columns were cut.
- Verify one calculated figure against the workbook, to catch stale cached values.
- Check the date columns actually show dates.
- Read the long labels in narrow columns — where the next cell is not empty, text is shortened to fit and ends in an ellipsis.