Every PDF-to-Word conversion is a reconstruction, and it is worth knowing that before you judge the result. The PDF does not contain paragraphs, columns, or tables — it contains characters at coordinates. A converter has to look at those positions and infer that these characters form a word, that these words form a paragraph, that this grid of positioned text was once a table.
Sometimes it infers correctly. On a complex layout it does not, and no converter will. Knowing that in advance turns "this tool is broken" into "this document was always going to convert badly, so I should take a different route."
Check what kind of PDF you have first
This determines everything.
Digitally created — exported from Word, InDesign, LaTeX, or a reporting system. It contains real text objects. Conversion has something to work with and results are often good.
Scanned — a photograph of paper. There is no text at all, so there is nothing to convert. You must run OCR first, which recognises characters in the image and produces text that can then be converted. See how to make a PDF searchable.
To tell them apart: try to select a line of text with your cursor. If you can, it is digital. If your cursor draws a box over a picture, it is scanned.
The methods
Microsoft Word (built in). Word 2013 and later open PDFs directly: File → Open, pick the PDF, confirm the prompt. Word runs its own reconstruction and gives you an editable document. This is the best free option for most people, it runs locally so nothing is uploaded, and quality is competitive with paid converters on straightforward documents.
Google Docs. Upload to Drive, right-click → Open with → Google Docs. Convenient with no software required, and it handles OCR on scanned files automatically, which is genuinely useful. Formatting fidelity is noticeably weaker than Word's — expect to rebuild the layout. Note that this uploads your document to Google.
LibreOffice. Free and cross-platform. LibreOffice Writer opens PDFs, though Draw usually gives better results for layout-heavy files. Runs locally, which matters for confidential documents.
Adobe Acrobat. The best conversion quality available, unsurprisingly, since Adobe defined the format. Requires a subscription, though the free trial covers a one-off job.
Online converters — Smallpdf, iLovePDF, Sejda, Adobe's web tool. Fast and easy, with free-tier limits. All upload your file to their servers; avoid for anything sensitive, and check retention policies if you use them for business documents.
What survives, and what does not
Setting expectations correctly saves an hour of frustration.
Usually fine: body text, basic bold and italic, simple bullet lists, single-column layouts, page breaks.
Usually damaged: multi-column layouts, which often interleave into nonsense; tables, which arrive as text boxes or lose their structure; headers and footers, which land as floating frames in the body; text boxes and callouts; footnotes, which detach from their references; and anything positioned precisely.
Almost always lost: the exact typography. Fonts that were embedded but are not installed on your machine get substituted, and substituted fonts have different metrics, so line breaks shift and pages reflow. A document that was four pages becomes five, and everything after the first table moves.
Frequently mangled: forms, which convert to static text; equations, unless they came from a source Word understands; vector diagrams; and hyperlinks.
When not to convert at all
When you only need a small change. If you are fixing a name, adding a date, or signing, converting to Word and back is a long way round — and it will reflow the entire document to fix three characters. Edit the PDF directly instead: Docento.app adds text, signatures, and annotations in your browser without uploading the file. See how to edit PDF text without Acrobat.
When the source file exists. Ask whoever sent the PDF for the original. Thirty seconds of asking beats an hour of cleanup, and people usually say yes.
When the layout is complex. For a heavily designed document, retyping the sections you need into a clean file is genuinely faster than repairing a broken conversion. Fighting a mangled multi-column layout can take longer than rebuilding it.
When the document must stay authoritative. A signed contract converted to Word is no longer the signed contract. Keep the PDF as the record.
Cleaning up afterwards
If you do convert, work through it in this order rather than fixing things as you spot them:
- Turn on formatting marks (¶) to see what actually happened. Converters produce a lot of manual line breaks, empty paragraphs, and text frames that are invisible until you look.
- Fix the paragraphs first. Converted documents typically have a break at the end of every visual line rather than every paragraph, which makes all subsequent editing miserable. Fixing this first is the highest-value step.
- Reapply styles. Converters produce direct formatting rather than heading styles. Applying real Heading 1/2/3 restores navigation, the table of contents, and — importantly — accessibility structure when you export back to PDF.
- Rebuild tables rather than repairing them. Almost always faster.
- Delete stray text boxes left over from headers and footers.
- Proofread OCR output specifically. OCR errors are plausible-looking substitutions —
rnform,1forl,0forO— that spellcheck often accepts. Numbers are the dangerous case, because a misread digit in a figure is invisible and consequential.
Going back to PDF
Once edited, export back with File → Save As → PDF in Word or File → Download → PDF in Google Docs. Use the export function, not Print to PDF — printing discards hyperlinks, bookmarks, and accessibility tags. See PDF vs Word for why PDF remains the right format for distribution.
