Why does my PDF show “Microsoft Word - Draft3” as its title?
You sent someone a PDF called 2026-annual-review.pdf. They opened it, and the browser tab said
Microsoft Word - Draft3.doc. Now you have looked at your own copy, and the tab says the same
thing on your machine too.
The file name on disk is fine. You can rename it a dozen more times and the tab will not change, because the string in the tab is not coming from the file name. It is stored inside the document, it has probably been there since long before this version existed, and it goes wherever the file goes — into the attachment, into the shared drive, into Google's search results.
Here is where it came from and how to get rid of it properly.
Two names, and your reader is showing you the other one
A PDF carries a small dictionary of information about itself, separate from anything on any page. One of its fields is Title. That is a different thing from the file name:
- The file name belongs to your disk. It changes whenever anybody drags the file somewhere or saves it with a new name, and it is not stored inside the document at all.
- The title belongs to the document. It was written once, usually automatically, and it travels with the bytes.
Which one you see depends entirely on the software. Chrome, Edge and Firefox's built-in viewers use
the title for the tab when the file has one, and fall back to the file name when it does not. Google
prints the title as the blue link text for a PDF result. Document management systems and e-filing
portals index it. Acrobat is the odd one out: it has a per-document setting under Document
properties → Initial view → Show, which is stored in the file as a flag called DisplayDocTitle,
and it usually defaults to showing the file name. That is why a file that looks perfectly respectable
on your desktop is embarrassing in someone else's browser tab.
Where "Microsoft Word - Draft3" actually comes from
There are two routes, and the exact string tells you which one you are looking at.
If it starts with "Microsoft Word - ", it was printed. When Word prints, the Windows print
spooler names the job Microsoft Word - <whatever the document is called>. Every print-to-PDF driver
— Adobe PDF, Microsoft Print to PDF, PDFCreator, CutePDF — takes the print job's name and writes it
into the PDF as the title, because from the driver's point of view that is the only name it has been
given. So the title is a snapshot of the Word file name at the moment somebody hit print, extension
and all. Draft3.doc means someone printed a file called Draft3.doc, probably around 2014.
If there is no prefix — just "Draft3", or "Template - final v2" — it was exported. Word's own Save as PDF / Export copies the Office Title document property (File → Info → Properties) into the PDF. Almost nobody ever fills that in deliberately; it gets set once, then inherited. Save a copy of an old draft and the property comes with it. Start from a departmental template and you inherit the template author's title. Rename the .docx forty times and the property does not move, because it is not the file name either.
Both routes explain the thing that makes this so annoying: the title is old. It records a moment in the document's history that everyone involved has forgotten, and nothing in a normal editing workflow ever revisits it.
Check what your file says before you change it
Our PDF inspector reads the title and author straight out of the file in your browser, without uploading anything. That is the fastest way to confirm the string is really in the document rather than something your reader invented.
If you have poppler installed, pdfinfo file.pdf prints the same fields in one line each. And
exiftool -a -G1 file.pdf is the one worth knowing, because it shows you the title twice — once as
PDF:Title from the info dictionary, and once as XMP-dc:Title if the file also carries an XMP
metadata packet. When those two disagree, you have found why an earlier attempt at fixing this
failed.
Setting it to something sensible
- Decide what the title should be. Not the file name — the name of the document as a human would
say it. "Annual Review 2025" rather than
2025-annual-review-final-v3. This is the string Google will print, so write it for a stranger. - Put the file into Edit PDF metadata and fill in only the Title box. Every box left empty is left alone, so correcting the title will not blank the author or the subject.
- While you are there, look at Author. It is filled in automatically from whoever's copy of Word made the file, and it is wrong roughly as often as the title is.
- Download it and check the tab. Open the new file in a browser. The notes also tell you if the document carried a second copy of these fields in an XMP block — that block is removed when you set anything, so the file has exactly one answer rather than two that can disagree.
Why the old one may still show up afterwards
You fixed it, you reopened it, and the tab still says Draft3. Three ordinary reasons, in order of likelihood:
- Your browser cached the file. Force a reload, or open the download from disk rather than from the URL you had open before.
- Google has not recrawled it. The title in search results is whatever was in the file when the crawler last fetched it. That can be weeks, and it is the same for a SharePoint or DMS index.
- The file had two titles and your tool only changed one. That is the classic case, it is common enough to have earned its own page, and we wrote it down separately.
If you also want the file name to match
The two are independent, which is usually what you want — a document called "Annual Review 2025"
filed as 2026-03-11 annual review.pdf is a sensible arrangement. When you do want both to say the
same thing, Rename PDF sets the file name and, by default, writes the same string in
as the title. Turn that option off and you get your original bytes back untouched, which matters if
the document is signed.
What fixing the title does not do
It changes nothing on any page. Not one word of the document moves. If the drafting history is also visible in the text, in a header, or in a comment, this does not touch it.
It is not anonymisation. Clearing the metadata panel leaves names in the page content, in annotations, in attachments and inside images, and people get this wrong in ways that matter — the full answer is here.
It does not change "Created with" and "Produced by". Those two fields will still name Word and whichever driver produced the file, and they say more about you than the title does. Clear them in the same run if that matters.
It does not set the Acrobat display flag. We write the title; we do not touch DisplayDocTitle,
so Acrobat may carry on showing the file name in its window bar. Browsers and search engines, which
are where the damage happens, read the title regardless.
The file is written out again, so a digital signature will no longer verify. Set the title before anything is signed.
Doing it at the source, and doing it in bulk
If this keeps happening, fix the document rather than the export. In Word, File → Info → Properties → Title — set it once and every PDF you export from that file inherits it. If you go via a print-to-PDF driver instead, the title comes from the file name at print time, so name the .docx properly before you print it.
For a folder full of files, three free tools beat any website:
exiftool -Title="Annual Review 2025" file.pdfworks on whole directories and is the quickest route. One warning that costs people an afternoon: exiftool writes PDFs as an incremental update, so the old title is still inside the file and can be reverted with-PDF-update:all=. That is fine for tidying, and not fine if the point was that nobody should see the old string.qpdf in.pdf out.pdfafterwards rewrites the document from its current objects and leaves the superseded ones behind, which is how you make an exiftool edit final.pdftk in.pdf update_info_utf8 info.txt output out.pdftakes a plain text file of InfoKey / InfoValue pairs, which is the sane way to script a hundred documents with different titles.
Draft3 was a real file once. It does not need to outlive everyone who worked on it.