# Keep formatting when converting PDF to Word Canonical HTML: [https://pdfpix.com/guides/keep-formatting-when-converting-pdf-to-word/](https://pdfpix.com/guides/keep-formatting-when-converting-pdf-to-word/) A PDF stores glyphs at coordinates and nothing about paragraphs, columns or cells, so every converter rebuilds the structure from measurements. Some structures map cleanly to Word and some never will. To convert PDF to Word without losing formatting, know which is which, use a converter that reads tags when the PDF has them, and check the result against the original with a comparison rather than a glance. ## Steps 1. **Check whether the PDF is tagged.** Open Document Properties in your PDF reader. Tagged PDF: Yes means the file carries paragraph and heading structure, and converts far better than an untagged one. 2. **Convert in this tab.** Open PDF to Word, choose the PDF, and read the count it reports after it reads the .docx back. A count well short of the PDF's text means a page did not convert. 3. **Fix structure before appearance.** In Word, turn text boxes back into paragraphs, rejoin split tables and reapply list styles. Then adjust fonts and spacing. 4. **Compare with the original.** Export the Word file to PDF and put both files into Compare PDF. The marked differences are the places still to fix. ## What maps to Word cleanly Body text in one column, with consistent line spacing, becomes paragraphs. Bold, italic and underline runs survive. Font names and sizes come through when the font is installed. Images placed in the flow become inline pictures. Page breaks land where the PDF's pages ended. Headers and footers are recognised when they repeat on every page. A simple table with ruled lines and one value per cell usually becomes a Word table. The general conversion steps, and the test for whether a PDF has text at all, are in [Convert PDF to Word](https://pdfpix.com/guides/convert-pdf-to-word/). This guide assumes the PDF has a text layer and asks how much of its layout can be kept. ## What a converter has to reconstruct Columns are the first casualty. A two-column page is two blocks of lines at different x positions, and a converter either builds Word columns, which reflow differently, or a text box for each column, which does not reflow at all. Bulleted lists are lines that begin with a bullet glyph; a converter that misses the glyph produces paragraphs with a stray character. Table cells without ruled lines are inferred from alignment, and a wrapped cell breaks the grid. Hyphenated line ends, drop caps, text on a curve, and footnotes with superscript markers each have no Word equivalent that a converter can be sure of. Expect to touch them by hand. A page that mixes all of them, such as a magazine spread or an annual report, is the case to budget an hour of cleanup for, whichever converter runs first. ## Tagged PDFs convert better A PDF exported from Word, Google Docs or InDesign with accessibility on carries a tag tree that names each paragraph, heading, list and table cell. A converter that reads the tags does not have to guess the structure. Word's own converter uses them. A PDF from a print driver, or one that has been through a compressor that strips structure, has no tags and gets the geometric reconstruction. The Tagged PDF line in Document Properties tells you which you have before you start. When the source document is still available, exporting a fresh tagged PDF from it costs a minute and removes most of the cleanup. ## PDF to Word keep formatting: the browser tool [PDF to Word](https://pdfpix.com/pdf-to-word/) converts a native-text PDF to a .docx in this browser tab. It reads every drawn run with its position and size, rebuilds paragraphs from that geometry, writes the .docx, then reads the .docx back and counts its text against the text the PDF held before it offers the download. The comparison is of words, not of pixels: it proves that no paragraph was lost or duplicated, which is the failure a visual check misses. It does not promise a pixel-identical document, because no converter can. The PDF stays in this tab throughout and no account is needed. A scanned PDF has no text layer, so the tool reports zero text rather than producing a document of pictures; that case needs an OCR step in a desktop app or a scanner app first. ## Fixing the result in Word Work on structure first. Open the Navigation pane; if it is empty, apply Heading styles to the headings, which also fixes the table of contents later. Select a stack of text boxes, cut, and paste as text into the flow. Convert stray bullet characters back to a list with the Bullets button. Rejoin a table split across a page break by deleting the paragraph mark between the two halves. Then appearance: set the font on the whole document once, set paragraph spacing once, and let Word reflow. Trying to match the PDF line for line is the way to spend a day; matching its structure takes an hour and produces a document that behaves like one. ## Proving the layout matches Eyes miss a dropped line on page 14. Export the Word document back to PDF and put it beside the original in [Compare PDF](https://pdfpix.com/compare-pdf/), which renders both in this tab, overlays the pages and lists the text that differs. Anything it marks is either a change you made on purpose or a place the conversion lost something. Both files stay in the tab. When the goal is to edit rather than to rebuild, [Open a PDF in Word](https://pdfpix.com/guides/open-a-pdf-in-word/) is the shorter path, and small edits to text may not need Word at all. The other document guides are under [PDF, Word, Excel and PowerPoint help](https://pdfpix.com/guides/office/). ## Questions and answers ### Why does my PDF come into Word as text boxes? The page had columns or floating elements, and the converter kept positions rather than inferring a flow. Cut the boxes and paste as text into the document body. ### Can any converter keep the formatting exactly? No. A PDF has no paragraphs or cells to copy, so every converter reconstructs them. Tagged PDFs reconstruct well; untagged, multi-column PDFs do not. ### What is a tagged PDF and why does it matter here? A tagged PDF carries a structure tree naming headings, paragraphs, lists and cells. A converter that reads it does not have to reconstruct the layout from positions. ### How do I check nothing was lost in the conversion? Read the count PDF to Word reports, then export the Word file to PDF and compare it with the original in Compare PDF. The list of differing text is the list of places to look. ### Should I fix fonts or structure first? Structure. Headings, lists and tables decide how the document behaves. Fonts and spacing are one setting each once the structure is right. ### Why did two columns become one long block? The converter read the lines in geometric order across both columns. Cut the second column's text out of the block and paste it after the first column ends. ## Related pages - [PDF to Word](https://pdfpix.com/pdf-to-word/): Turn a native-text PDF into an editable Word document. The file stays in your browser. No upload, no account needed. - [Convert PDF to Word](https://pdfpix.com/guides/convert-pdf-to-word/): Turn a PDF back into an editable Word document, and know what the conversion cannot recover. Free, in your browser, with PDF to Word. - [Open a PDF in Word](https://pdfpix.com/guides/open-a-pdf-in-word/): Open a PDF in Word to edit it, and know what Word's own converter changes. Free, in your browser, with PDF to Word. - [Compare PDF files](https://pdfpix.com/compare-pdf/): Compare two PDF versions and find changed, inserted and removed pages. The file stays in your browser. No upload, no account needed. - [PDF, Word, Excel and PowerPoint help](https://pdfpix.com/guides/office/): Find the right guide for PDFs inside Excel, Word and other software. Every guide on the job, sorted by the problem you have.