How to convert a PDF to Word

Short answer: open Word, choose File → Open, and pick the PDF. Word 2013 and everything after it — Word 2016, 2019, 2021 and Microsoft 365 — rebuilds the file as an editable document. It costs nothing extra, it happens on your machine, and no site sees your file. If Word is not installed, open the PDF with Google Docs instead, which is free and works on a Chromebook or a phone.

The rest of this page is the part that actually decides your afternoon: what the conversion gets wrong, and why.

Every question about PDF-to-Word conversion splits on this one fact, so check it before you pick a tool. Open the PDF in any reader and try to select a sentence with the mouse.

• The text highlights. The PDF carries real characters. Any route below works, and the free ones work well.
• Nothing highlights, or the whole page selects as one block. It is a scan — a photograph of a document. You need OCR, which narrows the field to Google Docs or Acrobat Pro.

This is also why conversions disappoint people. A PDF has no concept of a paragraph. It is a display format: a list of instructions saying draw this glyph at this coordinate in this font at this size. Converting to Word means reverse-engineering paragraphs, headings, tables and columns out of scattered coordinates. Every tool guesses, and on a plain report every tool guesses well.

Free if you already pay for Office, local, and better than its reputation.

1. Open Word and choose File → Open → Browse.
2. Set the file-type filter to All Files if the PDF is greyed out, and select it.
3. Word warns that it will convert the PDF into an editable document and that the result may not look identical. Click OK.
4. Wait — a long document can take a minute — then File → Save As → Word Document (.docx). Until you do this you are editing an unsaved conversion, not a file.

Where it shines: reports, letters, contracts, theses, anything that is mostly running text with headings. Word rebuilds real paragraph styles, which means the result is genuinely editable rather than a picture with text glued on top.

Where it fails: scans, because it does not run OCR — you get a full-page image. Also two-column academic papers, which it tends to unravel into one long column, and heavily designed pages, which become a lattice of text boxes.

The only free route that does OCR, which makes it the answer for scanned documents and photographed pages.

1. Upload the PDF to Google Drive.
2. Right-click it and choose Open with → Google Docs.
3. Drive extracts the text — running OCR automatically if the pages are images — and opens it as a Docs file.
4. Then File → Download → Microsoft Word (.docx).

The trade is layout. Google Docs prioritises getting the words out, so complex tables flatten, columns merge and images land in roughly the right place rather than exactly. For a scan there is no competition at this price; for a clean text PDF, Route 1 keeps more structure. And the file does sit on Google’s servers, which matters for anything confidential.

The most accurate option, and the only paid one. In Acrobat Pro: File → Export To → Microsoft Word → Word Document. Click Settings first if the PDF is scanned or in another language — that is where you enable text recognition and set the OCR language.

Acrobat has the advantage of having written the format, and it shows on structured documents: tables survive as tables, headings map to Word heading styles, and columns stay in their columns more often than anywhere else. If you convert PDFs weekly it is worth the subscription. For one document it is not.

Two things to know before searching for a workaround. Acrobat Reader cannot export to Word — every Export button in the free Reader opens a subscription prompt. And Adobe’s web version converts on Adobe’s servers, so the file is uploaded; the desktop app converts locally. The same split exists in the other direction, which we cover in the guide to Acrobat and Word.

LibreOffice is free, runs on Windows, macOS and Linux, and never touches a server — the right answer when the document is sensitive and you have no Word licence. Be warned about what it does, though: opening a PDF in LibreOffice launches Draw, not Writer, and Draw treats each line as a separate text box.

You can then File → Save As → Word 2007-365 (.docx), but the result is a drawing canvas of boxes rather than flowing paragraphs. Editing one sentence reflows nothing. It is fine for lifting text out and for documents you will retype around; it is poor for continuing to write in.

Better use of LibreOffice: open the PDF in Draw, select all, copy, and paste into a blank Writer document as unformatted text. You lose the styling and keep the words in editable form — usually the faster path.

Dozens of sites convert PDF to Word in a browser tab. Almost all of them work the same way: your file is uploaded, converted on their servers by a licensed engine, and deleted after some stated interval. The conversion quality is often good — many license the same engines Acrobat uses.

The catch is the upload. Reconstructing a PDF into paragraphs takes real processing that browsers cannot yet do well on their own, so unlike a Word-to-PDF conversion, PDF-to-Word genuinely needs a server for good results. Read that as: a marketing brochure, fine. A signed contract, a payslip, a medical letter or a passport scan, not fine. For those, Route 1 or Route 4 keeps the file on your machine.

Fonts get substituted. A PDF embeds a subset of each typeface — only the glyphs the document uses — and a subset cannot be installed as a system font. Word matches by name against what you have and substitutes when it cannot. Different widths, different line breaks, different page count.

Columns collapse. Two-column layouts have no marker in the file saying “this is a column”. The converter infers it from horizontal gaps and often reads the page as one wide column with strange spacing. Academic PDFs are the worst case.

Tables become approximations. A PDF table is usually just lines drawn near text. Converters that reconstruct a real Word table are guessing at cell boundaries; ruled tables with visible borders survive best, borderless ones rarely do.

Headers, footers and footnotes wander. They are drawn at fixed coordinates with nothing marking them as running elements, so they arrive as ordinary paragraphs at the top or bottom of the body, often mid-sentence.

A practical rule: convert only when you must. If you can get the original .docx from whoever sent the PDF, that beats any converter on this page. Conversion is for when the original is gone.

Most people converting a PDF to Word are not switching formats. They are making a round trip: the only copy of a form, a certificate or a proposal is a PDF, three words need changing, and it has to go back out as a PDF because that is what the recipient expects.

The return leg is the easy half. Word’s File → Save As → PDF does it — click Options and tick Create bookmarks using: Headings if the document has a structure worth keeping. Without Word, or on a machine where you would rather not install anything, the converter on the front page takes the .docx and writes the PDF inside your browser tab: nothing uploaded, no signup, no watermark, up to 20 documents at once, and the text stays selectable and searchable in the output.