How to merge PDF files without breaking the result

Merging is the most-used PDF operation and the least risky, because of how it works under the hood.

What merging actually does

A merge does not re-render anything. Each source document's page objects — their content streams, embedded font subsets, images and vector paths — are copied into a new document, and a fresh page tree is written that lists them in order.

That has one important consequence: quality cannot degrade. There is no re-compression step, so a 300 DPI scan stays a 300 DPI scan and vector text stays vector text. If someone tells you merging made their file blurry, something else happened — usually a compression pass afterwards.

It also means file size is roughly additive. Ten 2 MB files produce a file near 20 MB, minus a little from shared structure. If that is too big for where it is going, compress after merging, not before, so you only make one quality decision. Our file size guide covers the trade-offs.

The three things that actually go wrong

1. Order

This causes more re-dos than everything else combined. When you select multiple files at once, the browser hands them over in the order the file dialog listed them — which depends on the current sort and can differ from what you saw. Always confirm the list in the merge tool before running it, and use the reorder controls rather than trusting the initial order.

For recurring work, name files so that alphabetical order is correct order: 01-cover.pdf, 02-summary.pdf, 10-appendix.pdf. Note the leading zeros — without them 10 sorts before 2.

2. Orientation

Rotation is a per-page property that travels with the page. Merge a document whose pages are flagged as rotated 90° and those pages will still be sideways in the output. Fix the source first with Rotate PDF, then merge. Rotating afterwards works too, but you then have to identify which page numbers came from which file.

3. Encryption

A password-protected PDF cannot have its pages copied out until it is decrypted. If a merge fails on one file, that is usually why. Use Unlock PDF on a document you have the password for, then merge.

Merging versus appending

People often use these interchangeably; the distinction matters when the documents are related.

Merging produces a new document with a new page tree. Cross-references inside each source — internal links, a table of contents pointing at page numbers — refer to positions that have now shifted. A contents page that said "Section 3, page 12" will still say page 12 while the section now starts at page 47.

Appending in a full PDF editor can preserve or rebuild bookmarks and outline structure. If your documents carry bookmark trees you rely on, check the output before discarding the sources.

For the common cases — invoices, scans, application attachments, chapter files — none of this matters, because the sources have no internal navigation.

What survives and what may not

Reliably preserved: page content, text and its selectability, embedded fonts, images at original quality, page dimensions and rotation.

Sometimes lost, depending on the source: bookmarks and outlines, document-level JavaScript, form field values in complex AcroForm documents, some annotation types, and attachments embedded in the source file.

If your sources are interactive forms, fill and flatten them before merging rather than after.

A clean merge workflow

1. Check every source opens without a password.
2. Fix any wrong page rotation at source.
3. Remove cover sheets, blank scanner pages and duplicates with Remove Pages — it is easier to do this per file than to hunt through a 200-page merged document.
4. Merge, confirming the order in the list first.
5. Open the result and check the joins, not just the first page.
6. Compress only if the file needs to be smaller for its destination.

Privacy note

Merging on this site happens in your browser: the files are read into memory, the new document is assembled locally, and the result is handed back as a download. Nothing about the merge requires a server, which is why bundling a set of scanned identity documents does not involve sending them anywhere. The background on browser-based processing explains the mechanics and the limits.