A working link needs a link annotation: an invisible rectangle over the text with a destination attached. The blue underline is just styling.
Conversions that produce a visual copy of a page, particularly a virtual printer such as Microsoft Print to PDF, keep the styling and discard the annotations. Export from the source application instead.
You export a document full of references, send it out, and someone reports that none of the links work. Opening it yourself confirms it: the text looks exactly like a link and the cursor stays an arrow.
Appearance and function are separate
In a web page a link is one thing: markup that carries both the destination and the styling. A PDF splits those completely.
The blue underlined text is drawn onto the page like any other text that happens to be blue and underlined. It has no destination and no idea it looks like a link. The clickable part is a separate object, a link annotation, which defines a rectangle on the page and what should happen when someone clicks inside it.
Blue underlined text, drawn like any other text.
An invisible rectangle with a destination, stored separately.
Anything that captures appearance keeps the styling and drops the function.
Where links get lost
- Printed instead of exportedExport from the source
- Typed but never made a linkInsert as a hyperlink
- Document flattenedRe-export from the source
- Page rasterised to an imageRe-export from the source
- Internal link after mergingRenumbering broke it
- Destination no longer existsUpdate the target
The second case is worth mentioning because it catches people out. Typing a URL into a document does not always create a link. Some applications detect and convert it automatically, and others leave it as text that merely looks like an address. If the source document's link does not work when clicked there, it was never a link and the PDF is behaving correctly.
Checking whether links survived
Two quick methods, and both are worth running before sending anything link-heavy.
Hover before you click
Move the cursor over the link text. A real link annotation changes the cursor to a hand and usually shows the destination in a tooltip or status bar. Nothing happening means there is no annotation there.
Test one of each kind
An external web link, an internal cross-reference and an email address behave differently and fail independently. Checking one of each catches more than testing five web links.
Internal links after merging
A different failure with a different cause. An internal link, such as a contents entry, does not store a page number. It stores a reference to a specific page object.
Merging renumbers every object in the document. Tools that update those references keep the links working, and tools that do not leave them pointing at whatever now occupies that number, or at nothing. A merged report can therefore have a contents page whose entries jump to the wrong sections, which is worse than them not working at all because it looks like it is functioning.
PDFslime's Merge PDF carries external links and bookmarks across, and its own page says plainly that some internal page links do not survive any merge. What survives a merge more generally is covered in merging PDF files.
Fixing them
The reliable route is the same as for fonts: go back to the source and export properly.
Export or Save As PDF writes real link annotations from the links in the source document. It takes the same number of clicks as printing and produces a document where the links work, the headings become bookmarks, and the structure survives for anyone using a screen reader.
If the source is gone, links can be added to an existing PDF by drawing annotation rectangles over the relevant text and attaching destinations. It works and it is manual, so for a document with dozens of references it is worth some effort to find the original first.
One note on converters, ours included. PDFslime's Word, HTML and Markdown to PDF tools write a real link annotation for every link with a full web or email address, so those links work in any reader. A relative link or an in-page anchor has nothing to point at once the document is a standalone PDF, and stays as styled text. If your source leans on those, export it from the app that made it.
Merge files with the link boxes kept
Merge PDF carries external links and bookmarks across, in your browser. Test the internal links afterwards.
Common questions
Checking and restoring links.
01Why is my blue underlined text not clickable?
Because the colour and underline are styling drawn onto the page, while clickability comes from a separate link annotation. The styling survived and the annotation did not, which is why it looks like a link and behaves like ordinary text.
02Why does Print to PDF break links?
A virtual printer such as Microsoft Print to PDF only receives a description of how the page should look, and link destinations are not part of that. A browser's own Save as PDF is different: the browser writes the file itself, and Chrome's version keeps the links.
03How do I check whether my links work?
Hover over the text. A real link changes the cursor to a hand and usually shows the destination. Test one external link, one internal cross-reference and one email address, since they fail independently of each other.
04Why do my contents entries go to the wrong page?
Internal links reference specific page objects rather than page numbers. Merging renumbers every object, and tools that do not update those references leave links pointing at whatever now holds that number.
05Can I add links to a PDF that has none?
Yes, by drawing link annotations over the text and attaching destinations. It is manual work, so for a document with many links it is usually faster to find the source and export it properly.
06Do links work when the PDF is printed?
No. Paper has nothing to click, which is why documents likely to be printed should show important addresses in full rather than hiding them behind linked words.