Remove blank pages from a scanned PDF, and keep the ones that only look blank
By the getPDF team · Published 11 October 2026
The short answer
Drop the scan into Remove blank pages below, choose Scanned blank under Blank means, and press Remove blank pages: the empty backs of a double-sided scan are found and taken out, and the result lists the page numbers it removed. A page with real text is never removed, but a scanned back with only a signature or initials can count as blank, so read the list before you send the file. Your original stays untouched and nothing is uploaded.
Try it here, nothing is uploaded
Why is every other page blank?
The scanner read both sides of single-sided originals. Duplex mode is the right setting for a stack with print on both sides, and the wrong one for a stack of single-sided letters: every sheet gives a page of print and a page of nothing.
Two fixes, depending on whether you scan again:
- Next time, at the scanner. Scan single-sided stacks in single-sided mode, or switch on the scanner software’s own blank page setting. Epson Scan 2 calls it Skip Blank Page, with levels from Very Low (only truly blank pages) to High (pages with pen marks and show-through too); Epson’s support pages note you may need to try several levels (checked on 11 October 2026). Other makers have the same idea under other names.
- For the file you already have, the tool on this page.
How do I remove the blank pages, step by step?
- Drop the scan on the tool above. Several files, or a whole folder, work too: each is checked on its own.
- Pick what blank means under Blank means:
- Empty: only pages with no ink at all.
- Nearly empty: a few specks allowed. This is where the tool starts.
- Scanned blank: dust, punch holes and faint show-through allowed, as on the backs of a real scan.
- Press Remove blank pages. Each page is drawn small and its clearly dark pixels are counted; a page under the limit, and with no text on it, is taken out.
- Read the result. It says, for example, Removed 5 blank pages, then the file and the page numbers: removed pages 2, 4, 6, 8, 10.
- Download PDF. The file is named after yours with
-no-blanksadded. A file with nothing to remove comes back as it was, with a line that says so.
There is no preview grid to untick pages in. The check is the list of page numbers: if a number on it surprises you, look at that page in your original, which is still on your device, unchanged.
What counts as blank? What we measured
A scanned blank page is never pure white. There is paper grain, a grey cast, dust, punch holes and the text on the other side showing through. The tool counts only clearly dark pixels, so paper tint and light show-through do not count at all; specks, holes and pen strokes do.
To see where the 3 settings draw the line, we made a 13-page test scan: 6 printed pages, 6 backs of different kinds, and 1 typed page, at 150 dpi in colour. This is what each setting removed:
| Page | Empty | Nearly empty | Scanned blank |
|---|---|---|---|
| clean blank back (grain only) | removed | removed | removed |
| light show-through from the other side | removed | removed | removed |
| specks and 2 punch holes | kept | kept | removed |
| small initials, about 1 cm | kept | removed | removed |
| a signature line, about 6 cm | kept | kept | removed |
| strong show-through (dark mirrored text) | kept | kept | kept |
| typed: “This page is intentionally left blank.” | kept | kept | kept |
| the 6 printed pages | kept | kept | kept |
Read the bold cells twice. A back with initials or a signature is mostly empty paper, and to a pixel count it looks like a dusty blank. That is why the result lists every page it removed: on a stack of signed forms, check the list, or use Empty and remove the dusty backs by number.
The typed page shows the rule that protects digital documents: a page with real text on it is never removed, however little ink it has. A scan made searchable with OCR has a text layer, so remove blanks before you run OCR, not after; otherwise a back with show-through can carry a few OCR words and be kept. The guide on making a scan searchable covers OCR.
Speed: 120 scanned pages took about 3 seconds with Scanned blank in our test run on a mid-range laptop.
When the blanks are exactly every second page
If the scan is strictly print, blank, print, blank, you can also delete the even pages by number in Delete PDF pages: type 2, 4, 6, 8, 10 in Pages and press Delete pages. That is certain rather than measured, and it keeps the file’s bookmarks, taking out only those that pointed at a deleted page. It has no “even pages” shortcut, so on a 200-page scan typing 100 numbers is slower than letting Remove blank pages find them and checking the list.
For a handful of stray blanks in a long file, Organise PDF lists every page with a Remove button. A stack that also came out backwards can be fixed in the same pass with Reverse the order; see the backwards scan guide.
Delete pagesRemove pages, keep the outline. Free, runs on your device.The honest part: what it cannot tell apart
- Faint content on a scan. A page with 1 pencil note, a light stamp or a signature can read as blank, as the table shows. The tool never removes a page silently: the numbers are listed, and the original is not changed. Look before you send.
- Show-through that is dark. If the other side’s print shows through strongly, the back is not blank to a pixel count, and it stays. Run with Scanned blank, then delete what is left by number.
- Every page blank. If every page of a file looks blank, the tool stops and removes nothing, and suggests a stricter setting: the pages may hold very faint content.
- No preview or slider. The 3 settings are the control. If none fits your scanner, use the setting that keeps too many and delete the rest by number.
Everything runs in your browser: the scan is read, checked and saved on your device, and nothing is uploaded. Run the tool once, turn Wi-Fi off, and it still works.
Questions
Which setting should I use for a duplex scan?
Start with Scanned blank, which allows dust, punch holes and faint show-through. Then read the list of removed pages in the result. If a page with a signature or initials is on it, run the original again with Nearly empty, or delete the blanks by number.
Can it delete a page I need?
A page with real text, such as typed words or an OCR text layer, is never removed. A scanned page has no text, only a picture, so a back with a short pen note can count as blank. In our test, Scanned blank removed a page with a 6 cm signature line. Always check the list of removed pages.
Does it change my original file?
No. The result is a new file with -no-blanks added to the name, and the result card lists the page numbers it took out. Your original stays as it was, so you can run it again with another setting.
How do I stop the scanner making blank pages?
Scan single-sided originals in single-sided mode, or switch on the scanner software's blank page setting. Epson Scan 2, for example, calls it Skip Blank Page, with levels from Very Low to High.
Is there a page limit?
No. It runs on your device, page by page. A 120-page scan took about 3 seconds in our test run on a mid-range laptop.
The tools for this job
Keep reading
- The scanner fed the stack backwards: reverse PDF page orderReverse the page order of a PDF in 1 click when a scanner feeds the stack backwards.
- Merge odd and even pages from 2 scan passes into 1 PDF in reading orderScanned the fronts, then the backs? Interleave odd and even pages into one PDF in the right order, even when the backs are reversed.
- Delete PDF pages without losing bookmarks and linksDelete pages from a PDF and keep its bookmarks, links and form fields working.
- Make a scanned PDF searchable, without changing how it looksOCR a scanned PDF on your device: free, no account, nothing is uploaded.
- Organise PDF pages: merge, split, reorder and stamp, all in one placeMerge, split, reorder, rotate, delete and number PDF pages free, with no file limits.