Scanned PDFs are the worst offenders for file size. A ten-page text document exported from Word might be 200 KB; the same ten pages scanned can easily be 20 MB. The reason is simple: a scan isn't text at all — every page is a photograph, and photographs are heavy.
That also makes scans the hardest thing to compress well, because the images are the content. Squeeze too hard and the text turns to mush. This guide covers shrinking a scanned PDF properly, and the two steps most people skip that do more than compression alone.
Why your scan is so big
Three things drive the size:
- Resolution. Scanners often default to 300 or 600 DPI. That's wonderful for photographs and overkill for a page of text.
- Colour. A colour scan carries three channels of data per pixel. A black-and-white document scanned in colour is storing colour information about... white paper.
- Page count. Every page is a separate image, so size scales linearly. Fifty scanned pages is fifty photographs.
Knowing that tells you where the wins are — and compression is only one of them.
The right order: grayscale → compress
The single biggest mistake is jumping straight to the compressor. If your scan is in colour but the document isn't (a signed contract, a bank statement, a form), drop the colour first:
- Run it through Grayscale PDF — this alone removes two-thirds of the colour data.
- Then compress. The compressor now has far less to work with, so it can hit your target with much less aggression — which means the text stays legible.
Doing it the other way round (compress a colour scan hard) forces the compressor to destroy image quality to reach the same size. Same destination, much worse-looking document.
The steps
If it's a colour scan of a black-and-white document: convert it with Grayscale PDF first.
Open Blackpdf's Compress PDF tool and drop the file in.
Start with Recommended (Good Quality / Balanced) and look at the result. Scans respond dramatically to compression — Recommended often cuts a scan by most of its size while staying perfectly readable.
Only escalate if you must. Extreme (Maximum compression) will go further, but this is where scanned text starts to soften. Check the output before you send it — with a scan, always look at the result.
Click Compress PDF and download.
Where OCR fits (and where it doesn't)
A common misconception: OCR does not shrink your file. It adds an invisible text layer so the scan becomes searchable — useful, but it makes the file marginally bigger, not smaller.
The order that matters:
- Enhance → OCR → compress. Enhancing cleans the page (flat white background, sharper text), which makes OCR significantly more accurate. Compress last, so you're not degrading the image before the OCR engine reads it.
- Never OCR a heavily-compressed scan. Recognition accuracy drops on a mushy image. If you need searchable text, OCR before the aggressive compression, not after.
Common questions
My scan barely got smaller. Why?
Usually one of two things. Either it's already been compressed (running a compressor twice yields very little the second time), or the images are already at a low resolution and there's nothing left to take. If the file is still too big, the win is elsewhere: grayscale it, or remove pages you don't need.
The text went blurry after compressing.
You pushed too hard. Step back down from Extreme to Recommended, and grayscale the file first so the compressor doesn't have to work as hard to reach the same size. Blurry scanned text is almost always a sign of over-compression compensating for colour data that didn't need to be there.
Should I just re-scan it at a lower DPI?
If you still have the original document, yes — that's the cleanest fix. Scanning a text document at 200–300 DPI in grayscale produces a file that barely needs compressing. It beats rescuing a bloated 600 DPI colour scan.
Will compressing break the searchable text?
No. If the PDF has been OCR'd, the text layer is separate from the image and survives compression — you're shrinking the picture, not the text layer. But as above, don't run OCR after heavy compression, or accuracy suffers.
Can I compress a scanned PDF to a specific size?
Yes — on Pro, use Target size and enter your figure. See our target-size guides for what's realistic; scans are the case where very small targets (like 100 KB) get genuinely hard.
Wrap-up
For a scanned PDF, the order does most of the work:
- Grayscale it if the colour isn't carrying information.
- Compress with Recommended — check the result before escalating to Extreme.
- If you need searchable text, enhance → OCR → compress, in that order.
Scans are images, so they compress dramatically — the skill is taking the easy wins (colour, resolution, page count) before you start sacrificing legibility. For the general principles, see our full compression guide.