How to combine PDF files

The steps

Open the merge tool and choose two or more PDFs. Each file is read and its page count is shown, so you can see straight away whether the right documents are loaded. Drag the files into the order you want, press the merge button, and download one PDF.

The work happens in the browser tab. It is free, nothing is uploaded, no account is needed, and there is no limit on how many times you do it. For the common case of exactly two documents, the two file page is the same engine with less on the screen. To split a document instead of joining one, the organize tool cuts it into parts.

What survives the merge and what does not

Pages survive whole. Their text stays selectable, their page size stays what it was, and links inside a page keep working. A stack that mixes A4 and Letter stays mixed, because nothing is re-rendered to a common size.

Four things do not come across, because merging copies pages rather than the document structures around them: bookmarks, interactive form and signature fields, custom page labels such as roman numerals on a preface, and file attachments or named links. The tool reads each file before it runs and tells you which of your documents carry any of those, so the loss is never a surprise you find later.

The document properties are cleared as well. The merged file carries no title, author, producer or dates from any of the sources, which is usually what you want when the parts came from different people.

Order is the part that goes wrong

Set the order in the tool, not by renaming files first. Scanners, phones and mail clients name files in ways that sort badly: page 10 lands before page 2 in any plain alphabetical list, and a download named document (3).pdf tells you nothing about where it belongs.

Check the joins before you send it. Open the merged file and look at the first page of each part rather than reading the whole thing. One reversed pair is invisible to you and obvious to the person who receives it.

What the tool checks before it hands the file over

Every source page is read first and a receipt is taken of its text and its geometry. The merged file is then opened again and compared page by page against those receipts. The text of each page has to match exactly, the page count has to match, and every page has to keep its width and height.

If any of that fails the merge is refused rather than downloaded. That is the point of the check: a merged file with a silently dropped page is worse than no merged file, because nobody looks.

When merging is the wrong move

Two or three files that get read separately are better left separate. So is anything where one part gets replaced regularly, because you would rebuild the whole document every month.

If the parts are photos or screenshots rather than PDFs, they need to become pages first, and building one PDF per image and merging those is the long way round. The two file guide covers the quick case, and the merge hub lists the rest of the cluster, including scans, images and loose pages.