Scan a stack of single-sided paper on a duplex (two-sided) scanner and you get a PDF twice the size it should be — every real page followed by a blank one, because the scanner dutifully captured both sides of the sheet whether or not the back had anything on it. Combine PDFs from a few different sources and the same thing happens for a different reason: separator sheets, blank cover pages, an empty page left over from someone else’s scan job, all riding along in the final file. Either way, you end up with a document that’s bigger than it needs to be, awkward to print (nobody wants a blank page between every page of a contract), and looks unfinished when you send it to someone.
Going through page by page to find and delete each blank one works, but it’s tedious for anything longer than a few pages, and easy to miss one in a long stack.
How the detection works
Remove Blank Pages renders every page of your PDF and measures how much of it is actually covered in content versus plain white. A page where well under 1% of the pixels are non-white — which is what an empty scanned sheet looks like once you account for the faint speckling scanners tend to leave behind — gets pre-selected as likely blank.
It’s worth being precise about what this is and isn’t. It’s a pixel-coverage check, not OCR and not any kind of understanding of what’s on the page — it doesn’t read text, and it isn’t judging whether a page is meaningful, only whether there’s visibly anything there at all. That’s a deliberate, honest tradeoff: it’s fast, it runs entirely on your device, and it works well for the case it’s built for — a scanner-blank page with nothing on it. It also means a page with just a small stamp, a faint watermark, or a lone page number in the corner has real pixels on it and typically won’t be flagged, because there’s something there, however little.
There’s no sensitivity dial here, either — the threshold is fixed, and detection looks at the whole page rather than letting you exclude a header or footer strip from the check. That’s fine for a typical duplex-scanned stack, but if your originals are on letterhead — a logo or return address pre-printed on every sheet, including the ones that are otherwise empty — every one of those pages has real ink on it and will never come up as a candidate, no matter how sparse the rest of the page is. In that case detection just won’t help you; you’d select those pages by hand in the thumbnail grid the same way you’d correct any other miss.
Why it shows you the pages instead of just deleting them
This is not a blind auto-delete tool. After upload, you land on a page of thumbnails with the detected candidates already checked off — but nothing is removed yet. You look at the grid, and if the automatic pass got something wrong in either direction, you fix it yourself:
- Upload your PDF. Detection and thumbnail generation run automatically.
- Pages that look blank are pre-selected — you’ll see them checked in the thumbnail grid.
- Scroll through and adjust: uncheck anything that actually needs to stay (that faint stamp, a barely-visible watermark, a page you want kept for pagination even though it’s empty), and check off any blank page detection missed.
- Download once you’re happy with the selection — only then does anything actually get removed.
That review step matters more than it might seem. A heuristic based on how white a page looks will occasionally disagree with what you actually want kept — maybe a “blank” page carries a faint received stamp that’s evidence the document was filed, or a truly empty page is there on purpose to keep the pagination lining up for double-sided printing later. Because you’re looking at real thumbnails before committing, those cases stay in your hands instead of getting silently discarded.
Stopping blank pages before they happen
If you scan the same kind of single-sided paperwork often enough that this keeps coming up, it’s worth checking your scanner’s software for a blank-page-detection or blank-page-skip setting — most document scanners and multifunction printers with a duplex feeder have one, and turning it on means the blank backs never make it into the PDF in the first place. If yours doesn’t have that option, or you can’t find it, scanning in simplex (single-sided) mode instead of duplex does the same job for a stack that’s genuinely all single-sided: nothing but the side with content ever gets captured. Neither of those fixes a file that’s already sitting on your drive with blanks in it — that’s what the tool above is for — but it’s the difference between doing this cleanup every time you scan and not creating the problem to begin with.
Who this is for
- Anyone scanning single-sided paper on a duplex scanner — invoices, forms, signed pages — where every real page comes back paired with a blank one.
- Assembling a PDF from multiple scanned sources, where separator sheets or empty cover pages from different scan batches end up mixed into the final file.
- Cleaning up an old archive of scanned documents before sharing or filing them, so recipients aren’t paging through empty sheets to reach the content.
- Shrinking file size before emailing a long scanned document — blank pages still take up space, and cutting them is a quick, content-preserving way to trim a file down.
If your document doesn’t have obviously blank pages — say, you’re trying to remove a specific page that just happens to have very little on it, like a title page with one line of text — Delete PDF Pages gives you direct thumbnail selection without relying on any automatic detection at all.
And if what came back from the scanner is duplicate pages rather than blank ones — an auto-feeder that grabbed the same sheet twice, or fed one page through without advancing — that’s a different problem with a different shape, and Remove Duplicate Pages is built to catch that instead.
Everything stays on your device
Detection, thumbnail rendering, and the actual page removal all happen in your browser — your PDF is never uploaded anywhere. That matters here specifically because the documents that pick up blank pages from scanning are often the same ones you’d rather not hand to a server: signed contracts, filed forms, scanned paperwork with names and account numbers on it.