PDF to Word

How the structure is inferred

Every Word document you get from a PDF converter is a reconstruction. This one reads the position, size and font of each run of text, then rebuilds lines and paragraphs from the geometry.

• Text at the same height is joined into a line, and lines that sit close together with matching size and margin are joined into a paragraph, so you get flowing paragraphs instead of a hard break at the end of every line.
• A line that stops well short of the right margin is treated as the end of a paragraph, which is how most documents are set.
• Text noticeably larger than the body size becomes a Word heading at the matching level, so the navigation pane and table of contents work straight away.
• Bold and italic are read from the font names embedded in the PDF, and each run keeps its point size.
• Centred and right-aligned lines, such as titles and dates, keep their alignment.
• Lines that begin with a bullet or a number start a new paragraph rather than being merged into the one above.
• Links in the original remain clickable in the Word document, as long as they point to a web address or an email.

Where it struggles, and what to do instead

• A scanned PDF has no text at all, only a picture of it, so there is nothing to convert. Run it through OCR first, then convert the result.
• Multi-column layouts are read across the page rather than down each column, which interleaves the columns. For a two-column paper, expect to reorder text.
• Tables are not rebuilt as Word tables. Cells on one row come through on one line separated by tabs, which keeps the values in order and is quick to turn back into a table with Word's Convert Text to Table.
• Images are not carried over. Extract them separately and insert them where they belong.
• Headers, footers and page numbers are ordinary text in a PDF, so they appear in the document. Delete them, or use Word's own header feature.
• Text rotated at an angle, such as a DRAFT stamp or a margin note, is left out, because it would otherwise land in the middle of a sentence.
• A password-protected PDF has to be unlocked before it can be read.

Getting the best result

• The cleanest conversions come from PDFs that were exported from Word or another text editor in the first place. The structure was simple, so it is easy to recover.
• If you still have the original .docx the PDF was made from, use that. No conversion will beat the source.
• Turn on paragraph marks in Word after converting. Any stray breaks show up immediately and take seconds to remove.
• Fonts are mapped to their closest standard equivalent. If a PDF used a font you do not have, Word substitutes one and line lengths shift slightly.
• Check every figure in financial or legal documents against the PDF. Reconstruction cannot misread a number, but it can join two cells in an order you did not expect.
• For confidential files, this matters most: the conversion happens in your browser, so a contract or medical record never passes through someone else's server on its way to becoming editable.