Nobody merges PDFs because the operation is hard. They merge them because email, print shops and upload portals all want one file, and the source material arrives in pieces: a signed contract and two annexes, a chapter written in one tool and its figures exported from another, twelve scanned pages that belong to a single report. The awkward part is that PDF was designed as a self-contained document format, not as a page container you can concatenate, and that is why this job so often ends at a third-party site rather than in a browser tab. Merged files are also exactly the kind of document people hesitate to upload: a bundle of drafts, a bank statement, a signed agreement. This page links out to iLovePDF's merge page; the guide below is about what merging actually does to a file and when you should not let anybody's server near your documents.
Why you cannot just glue the bytes together
A PDF is a container with a header, a set of numbered objects, a cross-reference table or stream that says where each object lives, and a trailer pointing at the root catalogue. Every file carries its own numbering, its own offsets and its own trailer. Concatenate two files and you get two documents stapled together, with two object tables and two page trees. Some viewers are forgiving enough to render it; others report the file as damaged, and Acrobat will offer to repair it, which usually works and produces a messier file. This is the structural reason merging is not a browser-local job in the way image compression is: getting it right means renumbering objects and rebuilding the page tree, and the tools that do it properly are qpdf, pdftk, PDFsam, or the page-level merge in macOS Preview. Ghostscript can also repair a broken file on the command line.
What survives the merge and what quietly does not
Page content survives: text, images, embedded fonts and the page geometry of each source document, which is why a merged file often contains pages of different sizes. Everything layered on top of the pages is where surprises live. The document outline is usually rebuilt from the first file only, so bookmarks pointing into later documents disappear. Interactive form fields are frequently flattened into their current values, which is fine for a record and useless if someone still has to fill the form. Comments and annotations are often dropped, and tagged-PDF accessibility structure can be lost, which is a real problem if the merged file has to meet an accessibility requirement. Digital signatures definitely break, and that one is not a matter of degree: a signature covers specific bytes, and merging rewrites the byte layout. Merge first, sign last.
Getting the order and the front matter right
Order is the part people get wrong, because dragging files into a sensible sequence is easier than remembering which one is the cover page. Two habits help. Add the cover page and a table of contents as a separate PDF and merge it in first, rather than editing the assembled file afterwards. And decide about page size before you start: mixing A4 pages with US Letter pages is technically fine and looks broken in print, so if the document will be printed, normalise the sizes first or export the sources at a single page size. If the pages are scanned images rather than real text, remember that the images are the file size, so compressing the source images before merging usually beats compressing the merged PDF afterwards.
Frequently asked questions
Is merging PDFs free?
The service this page links to has a free tier that is genuinely usable for normal documents, and it sets limits on how much you can do per task and per hour; those numbers change over time, so read the current note on their page. Bulk users are expected to create an account or use a desktop version, which is the normal way these services fund themselves.
Are my PDFs uploaded to a server?
Yes. This is the honest trade-off of a link-out tool: the files are sent to the provider so their software can process them, so their privacy policy and retention rules apply. If the documents are confidential, a signed, or under a data processing agreement, merge them locally instead - qpdf and pdftk are free command line tools that do the same job with no network at all.
Do bookmarks, form fields and comments survive?
Usually not. The outline is generally taken from the first document only, form fields are often flattened to their current values, and comments and annotations are frequently dropped. Page content, images and fonts survive. If the merged file still has to be filled in or still has to be signed, merge first and sign after.
Why did the merged file open in a repair dialog?
That usually means the source file was already damaged, or that the merge produced a structure the reader is strict about. Open each input separately first and repair it individually - with Ghostscript or the reader's own repair option - and then merge the repaired copies. Merging a broken file makes the problem harder to see, not easier.