How to convert Excel to PDF without losing columns

You export a spreadsheet to PDF, open it, and the last four columns are missing. Scroll down and there they are — dumped onto their own pages at the end, stripped of the row labels that made them mean anything. A report that made sense in Excel has become something nobody can read.

This is the single most common complaint about spreadsheet conversion, and it is entirely fixable once you know what is happening.

Why it happens

A spreadsheet has no pages. It is an infinite grid, and it is perfectly happy to be forty columns wide because your screen scrolls.

A PDF is made of pages, and a page has a fixed width.

When something converts a spreadsheet to PDF, it has to decide what to do about that mismatch. The default behaviour, in Excel and in most converters, is to fill the page with as many columns as fit, then start a new page with the next batch. Nothing is deleted — but the table has been cut into vertical strips, and a strip of a table is not a table.

It is worth noticing that this default made sense once. It comes from printing, where you would tape the strips together across a desk. Almost nobody does that now.

The setting that fixes it

The fix is a scaling option, usually called fit to width or fit to one page wide. It tells the converter: whatever the width of this sheet, shrink it horizontally until it fits across a single page.

The important detail — and the one people get nervous about — is that fitting the width does not mean squashing the height. Those are two separate settings, and conflating them is what produces those unreadable PDFs where a hundred rows have been crushed onto one sheet at four-point type.

What you want is:

• Fit to 1 page wide
• Fit to as many pages tall as it takes

Every column lands on the page. Rows keep flowing downward. The type gets a little smaller — enough to fit the width, no more.

On Excel to PDF this is on by default, because it is the right answer for almost every real spreadsheet. If you genuinely want the old strip-cutting behaviour, you can switch it off.

How small will the text get?

This is the fair question, and it depends entirely on how wide your sheet is.

If your sheet is so wide that fitting it makes the text unreadable, no setting will save you. At that point the honest options are to switch to landscape, use a larger paper size, or split the sheet into logical groups of columns and convert each one separately with the identifying columns repeated.

Things worth doing before you convert

Most of the pain here is avoidable in Excel before the file ever reaches a converter.

Set a print area. If your sheet has notes off to the right or a scratch calculation in column AZ, the converter has no way to know those are not part of the report. It measures the used range. Selecting the range you actually want and setting it as the print area removes the problem completely.

Delete genuinely empty columns. An empty column with a stray space in row 400 is not empty. It counts toward the width and it will squeeze everything else.

Narrow the columns that do not need width. A date column set to 30 characters wide is taking space from the description column that actually needs it. Auto-fit the columns first and the whole sheet gets narrower for free.

Repeat the header row. In Excel, Page Layout → Print Titles → Rows to repeat at top. Without it, a table that runs to page four has column headings only on page one, and everything after that is a wall of numbers.

Check the page breaks. Page Break Preview shows you exactly where the cuts will fall, before you convert rather than after.

Landscape, and when to use it

Rotating the page gives you roughly 40% more width — an A4 page is 210mm wide in portrait and 297mm in landscape. For a wide table that is often the difference between comfortable and cramped.

The trade-off is that you get less height, so a long table needs more pages. For a table that is wide and short, landscape is almost always right. For one that is wide and very long, fitting the width in portrait usually reads better.

Multiple sheets

A workbook with several sheets converts all of them, in order, with the page setup applied to each. That is usually what you want.

If it is not — if you only need the summary sheet — it is faster to copy that sheet into a new workbook and convert that, rather than converting everything and deleting pages afterwards.

What about CSV files?

A .csv is not really a spreadsheet. It is a text file with commas in it, and it has no column widths, no fonts and no formatting to preserve.

Handing one to a spreadsheet engine means accepting that engine's import guesses, which is where the familiar problems come from: columns at a fixed default width that clip long values, a comma assumed as the separator when the file actually uses semicolons, and a header row that looks like ordinary data.

CSV to PDF measures each column against its real contents instead, works out the separator by looking at which one produces a consistent column count, and repeats the header on every page. If your source is a .csv, use that rather than opening it in Excel first — you will get a better table with less work.

The same applies to OpenDocument files

If you work in LibreOffice or OpenOffice, ODS to PDF does the same job with the same fit-to-width default. The engine that does the conversion is LibreOffice itself, so .ods is the format it handles most faithfully of all.