Convert from PDF · guide
What PDF to Word does
PDF to Word turns a text-based PDF into an editable Word document (.docx) inside your browser. Each horizontal line of text goes back where it was drawn, at its own size, bold and italic, so the page looks like the PDF. Paragraphs, headings and tables are not rebuilt, and the result lists what was carried across and what was left out. The PDF is never uploaded.
How to convert a PDF to Word
- Choose a PDF Select one PDF. Files exported from Word, a browser or another program convert best; a scanned PDF has no text to convert, so run it through OCR PDF first.
- Pick how it should come out Leave Looks like the PDF selected: it puts each line where it was drawn and starts a new Word page for each PDF page. Editable text is listed too, but it currently gives the same file, so it will not reflow as you edit.
- Convert and check Select Convert to Word, read the notes on what was carried across and what was not, then save the .docx and compare it with the PDF, especially tables and anything printed sideways, before relying on it.
Worked example
Recover editable text from a letter whose source document is missing.
- Starting point
- Choose a text-based PDF and review the available conversion mode.
- What to do
- Convert and open the DOCX in a word processor. Check that every line, page break and picture is where the PDF had it before editing.
- Result
- Expect a DOCX that looks like the letter, with each line in its own frame, not the original authoring file.
If the result is not right
If text is absent or incomplete, test whether it is selectable in the source. OCR a scanned copy first, then check recognised names and numbers.
Limits worth knowing
- Decorative layouts and multi-column pages may need manual correction. Do not assume identical layout or unchanged values.
Questions and answers
Will converting PDF to Word keep my formatting?
It keeps how the page looks, not how the document is built. Each horizontal line of text is placed where it was drawn, at its own size, bold, italic and font family, and the pictures it can read are placed at their positions, so the Word file looks like the PDF. Editing one line does not reflow the others, because every line is its own frame, and paragraphs, headings, lists and tables are not rebuilt. Text colour, drawn shapes, table borders and links are not carried over, and text printed sideways, such as a rotated table or a vertical margin label, is left out. The notes on the result say so. The Editable text choice on the page currently gives this same file.
Can I convert a scanned PDF to Word?
Not directly, because a scanned page is a picture with no words in it. A PDF that is scanned throughout is refused: run it through OCR PDF, which adds the missing text layer, and convert that result. In a mixed document, such as a typed letter with scanned signature pages, the typed lines convert as usual and the picture on a scanned page is placed where it was drawn when it can be read, with no words on it to select or search. The notes say how many pages had no text layer.
Will a two-column page come out in the right order?
Each line stays where it was drawn and nothing is reordered. Where the converter finds two columns on a page, the lines of each column are placed separately, so the left column stays on the left and the right on the right. Every line is its own frame, so text does not flow from one column into the next as you edit. Check short pages, forms and tables against the PDF, because side-by-side text that is not found to be columns can end up on one line. If the notes say pages were read as columns, that does not change where lines are placed. The Editable text choice currently gives this same result.
What happens to page headers, footers and page numbers?
They are kept where they were printed. Looks like the PDF treats a header, footer or page number like any other line, so each is text in a frame on its page rather than Word's own header or footer, and a page number will not update if you add or remove pages. Word keeps real headers and footers in a header area that this converter does not write, so add live ones in Word with Insert, then Header or Page Number, and delete the printed ones.
Which font will the Word document use?
Each run of text keeps its own family, mapped to one Word has: Helvetica, Arial and Liberation Sans become Arial, Times-style faces become Times New Roman, Courier and monospaced faces become Courier New, and Georgia, Garamond, Calibri, Cambria, Verdana and Tahoma keep their names. Where the PDF names nothing recognisable, it falls back to Arial, Times New Roman or Courier New, depending on whether the text is sans-serif, serif or monospaced. Sizes, bold and italic are read from the PDF, so the look is close, but a substitute font can run a little wider or narrower than the original.
Why didn't my table come through as a Word table?
Because the converter does not build Word tables. Looks like the PDF carries a table across as lines of text, not as cells, so the words and figures are there but the grid is not a Word table, and its borders and background colour are not carried over. Check the spacing between columns against the PDF. If you need the figures to calculate with, PDF to Excel is the better tool: it finds the tables it can read with confidence and puts each value in its own cell.