How do I combine multiple scanned images into one PDF?

Put the pages in the correct order first, then use a tool that assembles images into a single document. Watch for alphabetical sorting problems — scan10 sorts before scan2 unless the numbers are zero-padded.

My scanned PDF is too large to email. What should I do?

Scan at 300 DPI rather than 600, convert to greyscale if the document is black ink on white, then compress. The DPI choice has the biggest effect and has to be made before scanning.

I scanned both sides separately and the pages are out of order. Is there a fix?

Yes. An alternate merge interleaves two files, taking one page from each in turn, and handles the reversed back-side order that single-sided feeders produce. It avoids reordering dozens of pages by hand.

Can I search the text in a scanned PDF?

Only after optical character recognition, because a scan is an image of text rather than text. OCR works well on clean 300 DPI scans of printed material and poorly on handwriting or faint originals. Straightening crooked pages first improves accuracy noticeably.

How to Combine Scanned Pages Into One PDF

Published 22 August 2026 · 6 min read

Scanners and phone scanning apps have a habit of producing exactly what you did not ask for: twelve separate files, named in a way that sorts wrongly, a few pages upside down, one crooked, and a total size too large to email. Getting from that to a single tidy PDF is a five-minute job once you do the steps in the right order.

Get the order right first

Do this before anything else, because fixing order later means redoing the work. The classic failure is alphabetical sorting: scan1.jpg, scan10.jpg, scan11.jpg, scan2.jpg. Ten pages sort before two.

If you are renaming, pad the numbers — scan01, scan02, scan10 — and everything sorts correctly everywhere. If you would rather not rename anything, use a tool that lets you drag pages into order after loading them, which is usually faster for a dozen pages.

Then combine

What you use depends on what came out of the scanner.

Images (JPG/PNG), one per page — scan images to PDF is built for this: it takes a set of photos or scans and assembles them into a single document with consistent page sizing.
Images, but you want control over layout — images to PDF lets you set page size and orientation explicitly.
Several small PDFs — merge PDFs, reordering by dragging before you combine.
A double-sided scan done in two passes — fronts in one file, backs in another, backs in reverse order. Alternate merge interleaves them correctly instead of you reordering 40 pages by hand.

That last one saves the most time and almost nobody knows it exists. Scanning a stack double-sided on a single-sided feeder gives you fronts 1,3,5… and backs 6,4,2. Interleaving is a solved problem — do not do it manually.

Clean it up

Scans arrive with predictable defects. Each has a direct fix:

Crooked pages — deskew straightens pages that went through the feeder at an angle. Worth doing before OCR, which is noticeably less accurate on tilted text.
Wrong rotation — rotate whole pages or just the ones that came out sideways.
Inconsistent page sizes — fix page size normalises everything to A4 or Letter so the document prints predictably.
Blank backs — if you scanned double-sided and half the pages are empty, delete pages removes them in one pass.
Margins full of scanner bed — crop the document to trim the black edges.

Make it small enough to send

This is where most scanned PDFs fail. A dozen colour scans at 600 DPI produces an enormous file, because each page is a photograph rather than text.

Scan at 300 DPI, not 600. For text documents 300 is the standard, and it roughly quarters the data compared with 600. This is the single biggest lever and it happens before you start.
Convert to greyscale. A black-ink document does not need colour. Greyscale conversion typically cuts the size substantially with no meaningful loss.
Compress. Compress the PDF to reduce image data further. Check the result is still readable — over-compressed scans become mushy exactly where the small print is.

Make the text searchable

A scan is a picture of text, so searching it finds nothing. Extracting the text runs optical character recognition and gives you the content as text you can search, quote or paste.

Set expectations honestly: browser-based OCR is good on clean, straight, 300 DPI scans of printed text, and poor on handwriting, faint carbon copies or heavily skewed pages. Deskewing first genuinely helps. For a large archive of poor-quality scans, dedicated server-side OCR will do better.

Why do this locally

People scan documents for a reason, and the reason is usually official: contracts, certificates, medical letters, tax paperwork, identity documents. That is the exact category worth keeping off other people's servers. Every tool linked here runs in your browser.

Try it yourself

Merge PDFs — runs in your browser, nothing is uploaded.

Frequently asked questions

How do I combine multiple scanned images into one PDF?
Put the pages in the correct order first, then use a tool that assembles images into a single document. Watch for alphabetical sorting problems — scan10 sorts before scan2 unless the numbers are zero-padded.
My scanned PDF is too large to email. What should I do?
Scan at 300 DPI rather than 600, convert to greyscale if the document is black ink on white, then compress. The DPI choice has the biggest effect and has to be made before scanning.
I scanned both sides separately and the pages are out of order. Is there a fix?
Yes. An alternate merge interleaves two files, taking one page from each in turn, and handles the reversed back-side order that single-sided feeders produce. It avoids reordering dozens of pages by hand.
Can I search the text in a scanned PDF?
Only after optical character recognition, because a scan is an image of text rather than text. OCR works well on clean 300 DPI scans of printed material and poorly on handwriting or faint originals. Straightening crooked pages first improves accuracy noticeably.